Eview: Alternative Polyadenylation: A Twist On mRNA 3 End Formation
Eview: Alternative Polyadenylation: A Twist On mRNA 3 End Formation
Eview: Alternative Polyadenylation: A Twist On mRNA 3 End Formation
R
NA is the intermediary in the information trans-
fer process, standing between the genetic mate- A B S T R A C T Regulation of gene expression by RNA processing mechanisms is
rial, DNA, and the end product, the translated now understood to be an important level of control in mammalian cells. Regula-
protein. Messenger RNA transcripts (mRNAs) in eukary- tion at the level of RNA transcription, splicing, polyadenylation, nucleo-cytoplasmic
otic cells have, with few exceptions, similar architecture transport, and translation into polypeptides has been well-studied. Alternative
and features. Most mRNAs have 5= and 3= untranslated RNA processing events, such as alternative splicing, also have been recognized
regions (UTRs), exons, and introns, usually thought of as as key contributors to the complexity of mammalian gene expression. Pre-
“coding” and “noncoding” regions, respectively, a spe- messenger RNAs (pre-mRNAs) may be polyadenylated in several different ways
cial modified base at the 5= end, called a “cap”, and a due to more than one polyadenylation signal, allowing a single gene to encode
number of signals present within the mRNA that signal multiple mRNA transcripts. However, alternative polyadenylation has only recently
translational “start” and “stop” (Figure 1). taken the field as a major player in gene regulation. This review summarizes what
Almost all eukaryotic mRNAs acquire an uncoded is currently known about alternative polyadenylation. It covers results from bioin-
poly(A) tail at their 3= ends in a process called polyadenyl- formatics, as well as those from investigations of viral and tissue-specific studies
ation (reviewed in refs 1⫺5). This process involves two and, importantly, will set the stage for what is yet to come.
tightly coupled steps, in which the mRNA is first cleaved
at a specific site, and then adenosine (A) residues are
added in a nontemplated fashion. The process of 3= end
formation is interconnected to other transcriptional and
post-transcriptional processes, such as splicing and
transcriptional termination (reviewed in refs 6⫺8). De-
fects in 3= end formation can drastically affect the devel-
opment, growth, and viability of a cell.
The Mechanism of Polyadenylation. Virtually all
mammalian polyadenylation signals contain the con-
sensus sequence AAUAAA (or a variant) between 10 and
35 nucleotides upstream of the actual cleavage and poly-
adenylation site (Figure 2; ref 9 and references therein).
In addition, sequences 14⫺70 nucleotides downstream *Corresponding author,
of the polyadenylation signal (the U/GU rich binding [email protected].
site for the multisubunit cleavage stimulation factor, or
CstF) are known to be involved in directing polyadenyl-
ation (10−22). These two elements are often termed the Received for review June 10, 2008
“core” polyadenylation elements. Auxiliary elements and accepted August 21, 2008.
both upstream and downstream of the core elements Published online September 26, 2008
have also been characterized that can enhance poly- 10.1021/cb800138w CCC: $40.75
adenylation efficiency and have been identified in both © 2008 American Chemical Society
STOP
Figure 1. A schematic of a mammalian pre-mRNA. Blue boxes indicate exons, and A single series of processing events may take place
black lines indicate introns. AUG indicates the translational start codon for the open on a pre-mRNA in the simplest of scenarios. However,
reading frame, and the stop sign indicates the translational stop codon. AAUAAA
indicates a polyadenylation signal, and the red baseball cap is representative of the
it is becoming increasingly clear that this simplest sce-
monomethylated guanosine cap added at the 5= end of the mRNA. Lighter blue re- nario is often not the rule. Alternative processing, that is,
gions represent the 5= and 3= UTRs. choice of one or more distinct processing events, of
mRNAs is now recognized as an important regulatory
viral and cellular systems (23−41). Spacing of these el- level in gene expression control (reviewed in refs 9 and
ements relative to the core elements is significant in 47⫺49). In alternative splicing events, different mRNA
their influence on polyadenylation (24, 26, 27, 30, 41). transcripts could be generated from alternative use of
Often these auxiliary elements are polyadenylation effi- splicing signals. Alternative polyadenylation is defined
ciency elements; however, they may also provide addi- as use of more than one polyadenylation signal. Taken
tional functions in proper processing. together, alternative processing events could result in
In addition to the RNA sequence elements, five multi- RNA transcripts that could differ in subtle ways, such as
subunit protein factors make up the core mammalian having different length 3= UTRs, or in more dramatic
polyadenylation and cleavage machinery: cleavage and ways, such as encoding different proteins with different
domains. These alternative processing events can then,
polyadenylation specificity factor (CPSF), CstF, cleavage
in turn, influence transcript localization, stability, and
factors Im and IIm (CF Im and CF IIm), and poly(A) poly-
transport. Although alternative splicing is an incredibly
merase (Figure 2; refs 1⫺5).The polyadenylation ma-
important mechanism of generating transcript diversity
chinery assembles on the pre-mRNA before any reac-
and in regulation of gene expression (reviewed in refs 47
tions take place. Cleavage occurs 10⫺30 residues after
and 48), this review will focus on the lesser-known and
the AAUAAA; the nuclease responsible for this specific
previously under-appreciated mechanism of alternative
cleavage has been reported to be CPSF 73 (42). polyadenylation and, in particular, alternative polyadenyl-
Polyadenylation and 3= end formation play a critical ation in mammals.
role in gene expression because improperly processed It has only been recently appreciated that more than
mRNAs are not transported out of the nucleus into the half of the genes in the human genome are alternatively
cytoplasm to be translated. Likewise, inefficiently pro- polyadenylated (9). Therefore, a closer understanding
cessed mRNAs may be transported out of the nucleus of alternative polyadenylation as a mechanism for tran-
with lower frequency than those mRNAs that are effi- script diversity and gene regulation is warranted. The
ciently processed. Poly(A) tails have also been shown current review is directed primarily at highlighting dis-
to influence mRNA stability, and translation (reviewed coveries on mammalian alternative polyadenylation that
in refs 43⫺46). Polyadenylation, therefore, is a funda- have been made in the past several years. The many
mental aspect of gene expression. consequences of alternative polyadenylation will be dis-
cussed at the end of this review, as
well as emerging questions in the
CFI, CFII, PAP,
field. Several other reviews (4, 5, 47)
symplekin, etc.
have discussed aspects of polyadenyl-
CPSF CstF ation and alternative polyadenyl-
ation, including the implications of
polyadenylation on health and dis-
ease (50). In addition, recent studies
AAUAAA U/GU rich on alternative polyadenylation in
Auxiliary upstream Core upstream Core downstream Auxiliary downstream plants and algae, specifically rice,
element element element element Arabidopsis, and Chlamydomonas, is
Cleavage site
yielding fascinating results that may
Figure 2. A schematic of RNA sequences and protein factors involved in polyadenylation. Green boxes
have important ramifications for our
are core polyadenylation elements; light green boxes are potential auxiliary elements that may or may understanding of mammalian alterna-
not be present in a polyadenylation signal. Ovals are basal polyadenylation machinery as indicated. tive polyadenylation (51−53).
Type I STOP
AAUAAA
AAUAAA
and an ⬃3⫺4 kb mRNA is found in other tissues (73).
AUG
AAUAAA
AAUAAA
STOP
Type III STOP -adducin brain-specific alternatively polyadenylated
mRNAs, but this promoter is not found in humans (73).
The connection between cellular promoter choice and
polyadenylation signal choice has been further studied
Figure 3. Types of alternative polyadenylation. As described in ref 9, there are three in some other genes. Interestingly, a recent study sug-
basic types of alternative polyadenylation. Type I polyadenylation can be character- gests that tissue-specific expression of the X-linked
ized as constitutive polyadenylation, because only one polyadenylation signal is
gene MID1, involved in a malformation syndrome of
present in the 3= UTR, and is actually not alternative polyadenylation but is included
here for comparison. Type II alternative polyadenylation has more than one poly- midline structures called Opitz syndrome, is regulated
adenylation signal, but all polyadenylation signals are present in the 3= most exon. by concerted action of alternative promoters and alter-
Type III alternative polyadenylation involves alternative splicing coupled with native polyadenylation signals (74). In this study, it was
alternative polyadenylation. shown that cellular choice of a particular promoter dic-
were found in the brain (63) and in spermatogenesis tated which polyadenylation signal would define the 3=
(64). Finally, alternative polyadenylation patterns in end of the mRNA; however, a mechanism for how this
mammals seem to be conserved in evolution (56), sug- works is still unclear (74), although links between tran-
gesting selection pressure exists to keep these in place. scription initiation and transcription termination have
been discussed (7, 75).
These bioinformatic studies will provide an interesting
Finally, two recent studies report a connection be-
basis for validation in molecular biology laboratories in
tween tissue-specific alternative polyadenylation and
the near future.
translational efficiency (76, 77). A human apoptosis-
Tissue-specific polyadenylation of mRNA transcripts
inducing gene called hap was identified to have two dif-
has also been explored for a number of specific genes.
ferent sized mRNA transcripts, ⬃1.8 and ⬃2.7 kb, gen-
Although tissue-specific expression of some core poly-
erated by alternative polyadenylation (76). The larger of
adenylation factors themselves can and does occur
the two transcripts was found to be highest in brain and
(65−67), this has not yet been found to be a general
lowest in liver, while the ⬃1.8 kb transcript was high-
driving force in tissue-specific alternative polyadenyl-
est in testis and pancreas (76). Chloramphenicol acetyl-
ation. Here, I will mention a few examples of tissue- transferase (CAT) assays using reporter constructs con-
specific alternative polyadenylation of mRNA transcripts taining the polyadenylation signals from each of the two
from specific, noncore polyadenylation factor genes; transcripts showed that the ⬃2.7 kb transcript polyadenyl-
there are certainly many more that could be discussed. ation signal up-regulated translational activity ⬃3-fold,
Several examples are present in the recent literature while the overall levels of mRNA remained the same
demonstrating tissue-specific alternative splicing result- (76). Translational regulation coupled with alternative
ing in alternatively polyadenylated mRNAs (68−72). polyadenylation can have a negative effect as well.
These are all excellent examples of type III polyadenyl- BZW1 is a member of the bZIP superfamily of transcrip-
ation. Because of the recent nature of these studies, the tion factors, with a regular mRNA size of ⬃2.9 kb, but a
mechanism(s) of the alternative splicing (i.e., tissue- novel ⬃1.8 kb transcript was recently identified in the
specific expression of a regulatory splicing factor) have mouse testis (77). An AGUAAA suboptimal polyadenyl-
not yet been elucidated. ation signal is found 19 nucleotides downstream of the
A connection with transcription and alternative poly- stop codon, and use of this polyadenylation signal gives
adenylation has also been explored. The -adducin rise to the testis-specific transcript (77). Most interest-
gene gives rise to two transcripts, an ⬃8⫺9 kb mRNA ingly, reporter constructs using this suboptimal poly-
REFERENCES 6. Shatkin, A. J., and Manley, J. L. (2000) The ends of the affair: cap-
1. Wahle, E., and Kuhn, U. (1997) The mechanism of 3= cleavage and ping and polyadenylation, Nat. Struct. Biol. 7, 838–842.
polyadenylation of eukaryotic pre-mRNA, Prog. Nucl. Acids Res. 7. Maniatis, T., and Reed, R. (2002) An extensive network of
Mol. Biol. 57, 41–71. coupling among gene expression machines, Nature 416,
2. Keller, W., and Minvielle-Sebastia, L. (1997) A comparison of mam- 499–506.
malian and yeast pre-mRNA 3= end processing, Curr. Opin. Cell 8. Proudfoot, N. J, Furger, A., and Dye, M. J. (2002) Integrating mRNA
Biol. 9, 329–336. processing with transcription, Cell 108, 501–512.
3. Colgan, D. F., and Manley, J. L. (1997) Mechanism and regulation 9. Tian, B., Hu, H., Zhang, H., and Lutz, C. S. (2005) A large-scale anal-
of mRNA polyadenylation, Genes Dev. 11, 2755–2766. ysis of mRNA polyadenylation of human and mouse genes, Nu-
4. Zhao, J., Hyman, L., and Moore, C. (1999) Formation of mRNA 3= cleic Acids Res. 33, 201–212.
ends in eukaryotes: mechanism, regulation and interrelationships 10. Gil, A., and Proudfoot, N. J. (1984) A sequence downstream of
with other steps in mRNA synthesis, Microbiol. Mol. Biol. Rev. 63, AAUAAA is required for rabbit beta-globin mRNA 3= end forma-
405–445. tion, Nature 312, 473–474.
5. Edmonds, M. (2002) A history of poly(A) sequences: from forma- 11. Gil, A., and Proudfoot, N. J. (1987) Position-dependent sequence
tion to factors to function, Prog. Nucl. Acids Res. Mol. Biol. 71, elements downstream of AAUAAA are required for efficient rabbit
285–389. beta-globin mRNA formation, Cell 49, 399–406.