Regulation of gene expression in trypanosomatids: living with polycistronic transcription

Abstract

In trypanosomes, RNA polymerase II transcription is polycistronic and individual mRNAs are excised by trans-splicing and polyadenylation. The lack of individual gene transcription control is compensated by control of mRNA processing, translation and degradation. Although the basic mechanisms of mRNA decay and translation are evolutionarily conserved, there are also unique aspects, such as the existence of six cap-binding translation initiation factor homologues, a novel decapping enzyme and an mRNA stabilizing complex that is recruited by RNA-binding proteins. High-throughput analyses have identified nearly a hundred regulatory mRNA-binding proteins, making trypanosomes valuable as a model system to investigate post-transcriptional regulation.

Keywords: Leishmania; Trypanosoma; mRNA decay; mRNA processing; mRNA translation; transcription.

Figure 1.

Life cycles of T. brucei and Leishmania. These simplified diagrams show the main developmental stages mentioned in this review. Different surface molecules are colour coded. The different stages are named according to the presence or absence of an external flagellum, overall shape and the position of the kinetoplast relative to the nucleus. In trypomastigotes (dotted arrows), the kinetoplast and flagellar base are posterior to the nucleus; in promastigotes, they are anterior. Amastigotes are amotile and have only a flagellar stub. Non-dividing transmission forms are labelled in grey and the macrophage lumen is pale yellow; QS means quorum sensing. Transformation of bloodstream-form trypanosomes to procyclic forms can be induced in vitro by applying specific stimuli, such as cis-aconitate, and changing the culture conditions (providing proline, decreasing the temperature). (b) Transformation of Leishmania metacyclic promastigotes to axenic amastigote-like forms can be achieved by elevating the temperature and the medium acidity. Culture temperatures for amastigotes are species dependent.

Figure 2.

Gene expression in kinetoplastids. This diagram is modified from [11]. The thick line at the top represents the genome, and each coloured block is represented in a mature mRNA. The corresponding mRNAs are shown with the coding regions thicker than the untranslated regions. The numbered steps are described in the text. 1: Modified histones in an RNA polymerase II initiation region. 2: RNA polymerase II elongation. 3: RNA polymerase II termination. 4: Endonuclease cleavage of precursor (postulated). 5: Trans-splicing and polyadenylation. 6: Incompletely processed mRNAs can be degraded by the exosome. 7: Export of a completed mRNA, with bound poly(A) binding protein (PABP), exon junction complex (EJC) and nuclear cap-binding complex (CBC). (Note that we do not know which, if any, splice junctions are bound by the EJC, and also that mRNA export can commence before the 3′-end is complete.) 8: Emergence of a mature mRNA including proteins on the coding region (a) and a specific, stabilizing protein (b) bound to the 3′-untranslated region (3′-UTR). 9: Binding by a silencing or aggregating RNA-binding protein (c) and condensation into granules. 10: Binding of EIF4E, EIF4G and EIF4A, and translation. 11: Protein (b) is replaced by a destabilizing RNA-binding protein (d) and deadenylation starts. 12: Decapping by ALPH1. 13: Degradation by XRNA and the exosome. 14: Rapid decay pathway—immediate decapping promoted by protein (e). Many of the proteins involved are listed in electronic supplementary material, table S1.

Figure 3.

mRNA processing. (a) This shows the region between two different open reading frames, ORF1 and ORF2. The distance between the polypyrimidine tract and polyadenylation site in T. brucei is about 100 nt. The nature of the interaction between the polyadenylation and splicing complexes is unknown; hypothetical linking proteins are shown but the interaction might instead be direct. (b) The simplest possibility for processing the primary transcript in (a), when only a single splicing signal is present. (c) This is a precursor with two possible splicing signals, labelled as A and B. If neither signal is used, the mRNA will contain more than one open reading frame, and only protein (1) will be produced. (d) If both sites from precursor (c) are used, the RNA from ORF1 (i) has a short 3′-UTR, RNA (2) has a short 5′-UTR (ii) and there will be an additional processed RNA with no open reading frame from intergenic region (iii). This pattern has been documented in detail for the procyclin [101,102] mRNAs but is doubtless seen in many more [29]. (e) If only site A of precursor (c) is used, the RNA for ORF2 will have a long 5′-UTR (iv) that includes splicing signal (b). (f) If only the distal site B f precursor (c) is used, the ORF1 mRNA will have a long 3′-UTR (v) which includes polypyrimidine tract (a).

Figure 4.

Interactions of cap-binding translation initiation complexes in T. brucei and Leishmania. The key is on the bottom right. The colours of the ovals identify different EIF4E interaction partners. Text colour codes for ‘activator’ (magenta) and ‘repressor’ (cyan) indicate whether the protein activates, or represses, gene expression when tethered within the 3′-UTR of a reporter mRNA in bloodstream-form T. brucei [41,167] (electronic supplementary material, Note 30). The solid frames indicate that the protein could be UV-cross-linked to poly(A)+ RNA in bloodstream-form T. brucei [41]. Gene numbers are from T. brucei, with the prefix ‘Tb927.’ removed. For results that rely on pull-down only, arrows point away from the bait protein. Suggested stage-specific interactions between Leishmania EIF4E1 and both EIF3 and EIF4G3 [168] are not shown. Binding activities of Leishmania EIF4E4 to modified caps are reported in [169,170]. Alternative names are G1-IP2 for RBP43, and G1-IP for Tb927.11.6720. References for pull-down and interaction results are as follows: T. brucei PABPs [155]; Leishmania and T. brucei EIF4E1 [168,171]; T. brucei EIF4E2 [172]; Leishmania EIF4G3 [173]; T. brucei and Leishmania EIF4E3 and EIF4E4 [168,174]; T. brucei EIF4E5 [175]; T. brucei EIF4E6 [176]. The tethering results for EIF4E5 and EIF4E6 are unpublished (L. Melo do Nascimento and C. Clayton 2019, unpublished data); note that the activities of these proteins when bound to the 5′ cap may not be the same as those seen in the tethering assay.

Figure 5.

RNA-binding proteins that have been implicated in developmental regulation. The figure shows a simplified pathway from the bloodstream form to the metacyclic form. The proteins shown were chosen because they are well characterized. Several more proteins have been shown to be essential for differentiation; these, as well as other RNA-binding proteins that are expressed in a stage-specific manner, but whose roles in differentiation have yet to be defined in detail, are mentioned in the text or can be found in electronic supplementary material, table S1. Regulated expression is derived from proteome data for differentiation of bloodstream forms to procyclic forms [181,253,254] and for RBP6-induced differentiation of procyclic forms to metacyclic forms [255]. Proteome data for epimastigotes are not yet available. *RNAi targeting both simultaneously is lethal. **RBP6 expression triggers differentiation to epimastigotes, and the mRNA is present in proventricular and salivary gland parasites, but it is not known whether RBP6 is essential for differentiation. References for the phenotypic results are in the text. ‘Essential’ usually means that RNAi was detrimental; the absence of an RNAi effect does not necessarily mean that the protein is not required.

Figure 6.

Lengths of 3′- and 5′-UTRs for mRNAs encoding proteins belonging to different functional classes. Protein classes were assigned manually, according to genome annotations and publications [171]. UTR lengths were downloaded from TritrypDB and are shown on a log scale (electronic supplementary material, Note 40). The boxes indicate the 25th and 75th percentiles, and the bar within the box is the median (label is in nt). Whiskers extend to the most extreme measurement that is within 1.5× the inter-quartile range, and spots are outliers. The dark grey line indicates the overall median, and the pale grey area is between the overall 25th and 75th quartiles. Colours indicate classes in which the medians for the 5′- and/or 3′-UTRs are outside the overall 25th–75th percentile range.