Lecture 2. The odessey of mRNA: An appreciation of the trials and tribulations of RNA

Figure 1 provides a sketch of what a mRNA has to go through before it can perform its most obvious biological function of coding for a protein.

-Transcription (you heard about this in earlier lectures)

-As the RNA is transcribed it is "capped" on its 5' end. This only occurs with RNA polymerase II (pol II) transcripts. As suspected, the enzyme that does has been shown to associate with pol II -its CTD (C-terminal domain).

[There was prior evidence for this type of interaction for a viral polymerase and capping enzyme (Hagler & Shuman (1992) Science 255, 983-986: "A freeze-frame view of eukaryotic transcription during elongation and capping of nascent mRNA").]

-Nuclear pre-mRNAs from higher eukaryotes often contain more intron than exon sequence.

-The introns are removed in the nucleus by the spliceosome in a process that involves several steps (Aluded to in last lecture).

(Wahle & Keller, Annual Rev. Biochem. (1992) 61, 419-440: "The biochemistry of 3'-end cleavage and polyadenylation of messenger RNA"; Cogan & Manley, Genes & Devel. (1997) 11, 2755-2766: "Mechanism and regulation of mRNA polyadenylation".)

-The polyadenylation factors are also associated with the CTD of Pol II. (Hirose & Manley (1998) Nature 395, 93-96: "RNA polymerase is an essential mRNA polyadenylation factor.")

-Several examples of control of polyA addition have been identified.

The processed mRNA exits the nucleus via a nuclear pore.

The mRNA can associate with ribosomes and be translated to give protein.

-What’s wrong with this picture: Figure 1 revisited!!

IIA. RNA as a ribbon?

-RNA, left alone, folds into multiple structures that are stable for long periods of time:

-Isolated tRNAs often need to be renatured in order to be charged by tRNA amino acyl synthetases: Adams, Lindahl & Fresco, Proc. Natl. Acad. Sci. USA (1967) 57, 1684-1691: "Conformational differences between the biologically active and inactive forms of a tRNA").

-RNAs often adopt multiple and inactive structures. For example, often in self-splicing reactions, a fraction of the group I intron will excise itself fast, but the remaining fraction will self-splice much slower. Sometimes the RNA can be renatured (e.g. Walstrum & Uhlenbeck, Biochemistry (1990) 29, 10573-10576).

-Misfolded RNAs can be extremely stable kinetically, with alternative conformations often remaining for days (e.g., Fedor & Uhlenbeck, Proc. Natl. Acad. Sci. USA (1990) 87, 1688-1672: "Substrate sequence effects on "hammerhead" RNA catlaytic efficiency").

-20 amino acid side chains vs 4 RNA side chains: Diversity & uniqueness.

-In secondary structure, protein side chains point out, ready for interactions that can uniquely determine tertiary structure; RNA side chains point in in base pairs, masking their differences.

-Protein secondary structure is of marginal stability, so that if the wrong secondary structure forms it is quickly unformed; in contrast RNA secondary structure is very stable. For example, a duplex with 10 base pairs can take a year to dissociate into single strands!!

-Proteins can help RNA fold and unfold. For example, proteins can denature tRNA. hnRNP's and related proteins can speed duplex formation and dissociation (Pontius & Berg, Proc. Natl. Acad. Sci. (1990) 87, 8403-8407; Tsuchihashi, Khosla & Herschlag, Science (1993) 262, 99-102; Herschlag, Khosla, Tsuchihashi & Karpel, EMBO J. 13, 2913-2924).

-The problems that RNA has in folding and misfolding must be dealt with by the cell.

As the RNA transcript is made, it is coated with proteins to form hnRNP particles (heterogeneous nuclear ribonucleoprotein). There are >20 different hnRNP proteins that seem to coat the RNA (Dreyfuss, Matunis, Pinol-Roma & Burd (1993) Annu. Rev. Biochem. 62, 289-321: "hnRNP proteins and the biogenesis of mRNA.")

(Herschlag (1995) J. Biol. Chem.270, 20871-20874: "RNA Chaperones and the RNA Folding Problem"). (Chaperones are proteins that help proteins fold correctly by preventing their misfolding into aggregates.)

Relating back to the RNA world, it is reasonable (but not established) that the hnRNP's were initally used as RNA chaperones and that later in evolution they were co-opted into additional functions:

-To help distinguish different mRNAs, and perhaps control their turnover. hnRNP proteins have been shown to exhibit some specificity in the RNA sequences that they prefer (e.g., Bennett, Pinol-Roma, Staknis, Dreyfuss & Reed, Molec. Cell. Biol. (1992) 12, 3165-3175: "Differential binding of heterogeneous nuclear ribonucleoproteins to mRNA precursors prior to spliceosome assembly in vitro").

-To help splicing or help control splicing. hnRNP protein A1 has been shown to influence splice site choice in vitro (Mayeda & Krainer, Cell (1992) 68, 365-375: "Regulation of alternative splicing by hnRNP A1 and splicing factor SF2").

-To help retain RNAs in the nucleus until they are fully processed and to aid in transport out of the nucleus (see below).

Now we have hnRNP, but pol II keeps going after the cleavage event, continuing on to make apparently non-functional transcripts that are hundreds or thousands of nucleotides long. These RNAs are rapidly degraded, presumably because they lack the 5' cap.

-Only about 5% of total RNA makes it out of the nucleus into the cytoplasm!

-The lost RNA is from introns, the 3' region beyond the cleavage/polyadenylation site, and turnover of RNA without release.

-Why is the process so complex and so inefficient!?

1. For control

2. Because the evolutionary driving force is not large enough.

-This can arise if the organism doesn't care as much about energy efficiency as environmentally concerned individuals care about fuel efficiency in automobiles. Alternatively, the organism might care, but there may be no accessible selection pathway toward greater efficiency. (I.e., How facile are the evolutionary mechanisms to evolve greater efficiency? How easily can the molecules be rearranged through selection to do the job?)

-Compare with group II self-splicing: Same chemical pathway, but now >50 proteins & 5 RNP complexes!

-Why is pre-mRNA splicing so complex? (And for that matter, all of RNA processing)

1. It has a more difficult problem than in self-splicing, as many different exon/intron junctions must recognized. Especially in mammalian splicing, the consensus sequences are quite weak. It has recently been suggested that proofreading is involved to ensure choice of the correct exons (Guthrie, Science (1991) 253, 157-163: "Messenger RNA splicing in yeast: clues to why the spliceosome is a ribonucleoprotein"). Need to break a process up into distinct steps to allow proofreading.

The general principle of proofreading is that an input of energy is required to allow organization and control -there's no such thing as a free lunch. (It is also worth reading about and understanding the thermodynamic and kinetic requirements for proofreading as this comes up in all areas of biochemistry. [Thopmson (1988) TIBS 13, 91-93: "EFTu provides an internal kinetic standard for translational accuracy"; Yarus (1992) TIBS 17, 130-133: "Proofreading, NTPases and translation: constraints on accurate biochemistry"; the classic in the field, though incorrect in the detailed mechanism suggested, is Hopfield (1974) Proc. Natl. Acad. Sci. USA 71, 4135-4139: "Kinetic proofreading: a new mechanism for reducing errors in biosynthetic processes requiring high specificity"; Burgess & Guthrie (1993) TIBS 18, 381-384: "Beat the clock: Paradigms for NTPases in the maintenance of biological fidelity."]

2. A complex pathway can allow multiple options for control, allowing alternative splicing. Examples of alternative splicing abound. The classic example is sex determination in Drosophila. The basic principles here, at the very crude level they are known, appear to be analgous to those in trascriptional control: The use of protein factors, whose expression or activity is controlled, to either recruit a splicing factor to a weak splicing site or to block a strong splicing site.

Mattox, Ryner & Baker, J. Biol. Chem. 267, 19023-19026: "Autoregulation and mulifunctionality among trans-acting factors that regulate alternative pre-mRNA processing"

Valcarcel, Singh, Zamore & Green, Nature (1993) 362,171-175, "The protein Sex-lethal antagonizes the splicing factor U2AF to regulate alternative splicing of transformer pre-mRNA".

Tian & Maniatis, Science (1992) "Positive control of pre-mRNA splicing in vitro"

Berget (1995) J. Biol. Chem. 270, 2411-2414: "Exon recognition in vertebrate splicing"

Hertel, Lynch & Maniatis (1997) Curr. Opin. Cell Biol. 9, 350-357: "Common themes in the function of transcription and splicing enhancers"

Lopez (1998) Annu. Rev. Genet. 32, 279-305: "Alternative splicing of pre-mRNA: Developmental consequences and mechanisms of regulation"

Grabowski (1998) Cell 92, 709-712: "Splicing regulation in neurons: Tinkering with cell-specific control"

-Some of the details:

Six base pairing rearrangements in the 1st step (Figure 3).

Several steps require DEAD box proteins (Figure 4)

-What are DEAD box proteins and what do they do?

-Share sequence/ strucutral homology with DNA helicases

-RNA-dependent ATPases

-Some can unwind RNA duplexes

-Involved in every interesting biological process that utilizes RNA:

-Splicing

-mRNA transport

-Ribosome biogenesis

-Translation initiation

-RNA localization

-RNA turnover

-But in no case is the molecular process understood (Figure 5)

-Subset of hnRNP and SR proteins shuttle between nucleus & cytoplasm.

-Thus transport is presumably of (remodeled) protein/RNA complex.

-The 5'"cap" may serve as an export signal

-However, very little is known about how RNA is transported.

-hnRNP's and snRNP's appear to be rather specifically localized within the nucleus, and there is evidence that hnRNP's are associated with the nuclear matrix.

(Subcellular organelles should not be considered "bags of enzymes".)

-In vivo hybridization experiments have suggested that a viral RNA exits the nucleus on "tracks" that lead to a nearby nuclear pore. Such localized transport into the cytoplasm might help in establishing organization within the cytoplasm, but remains a controversial notion. (Lawrence, Singer & Marselle, Cell (1989) 57, 493-502: "Highly localized tracks of specific transcripts within interphase nuclei visualized by in situ hybridization")

-Must find translation initiation factors and ribosomes (next Lectures)

-Specific RNAs can be localized to certain regions within the cytoplasm. This is thought to play a role in establishing gradients that are necessary for development (Gottlieb, Curr. Opin Cell Biol. (1990) 2, 1080-1086: "Messenger RNA transport and localization"; Melton, Science (1991) 252, 234-241: "Pattern formation during animal development"; Mowry & Melton, Science (1992) 255, 991-994: "Vegetal mRNA localization directed by a 340-nt RNA sequence element in Xenopus Oocytes"; Sundell & Singer, Science (1991) 253, 1275-1277: "Requirement of microfilaments in sorting of actin mRNA")

-mRNA's in the cytoplasm vary in their lifetime, with the average half-life being ~30 min. Some sequences important in determining turnover rates have been. The turnover is controled in many cases, with factors both increasing and decreasing the turnover. (e.g., Malter & Hong, J. Biol. Chem. (1991) 266, 3167-3171: "A redox switch and phosphorylation are involved in the post-translational up-regulation of the adenosine-uridine binding factor by phorbol ester and ionophore"; Ratnasabapathy, Hwang & Williams, J. Biol. Chem. (1990) 265, 14040-14055: "The 3'-untranslated region of apolipoprotein II mRNA contains two independent domains that bind distinct cytosolic factors")

-RNA editing

(Bass "RNA editing: New uses for old players in the RNA world" , "The RNA World (1993) Cold Spring Harbor Press, pp 383-418; Simpson "RNA editing -An evolutionary perspective", in ‘The RNA World’ (1999) Cold Spring Harbor Press, pp585-608.)

-RNA modifications

-Nonsense mediated decay

-Nuclear amino acylation of tRNA

-RNAi

(Sharp (1999) Genes & Devel. 13, 139-141: "RNAi and douible strand RNA")

-Frameshifting in translation

-Protein splicing

(Perler et al. (1992) PNAS 89, 5577-5581: "Intervening sequences in an Archaea DNA polymerase gene"; Gimble & Thorner (1992) Nature 357, 301-306: " Homing of a DNA endonuclease gene by meiotic gene conversion in Sacchoromyces cerevisiae".)

-Look at all of the RNAs

Hypothesis-driven vs hypothesis-limited

- Pictures are appealing, but misleading by necessity of their static nature.

-Can single molecule detection in vivo save the day?

-Femino, Fay, Fogarty & Singer (1998) Science 280, 585-590: "Visualization of single RNA transcripts in situ".

-Arn & Macdonald (1998) Cell 95, 151-154: Motors driving mRNA localization: New insights from in vivo imaging".