Does protein synthesis occur in the nucleus?
James E Dahlberg1 and Elsebet Lund
Although it is universally accepted that protein synthesis occurs
in the cytoplasm, the possibility that translation can also take
place in the nucleus has been hotly debated. Reports have been
published claiming to demonstrate nuclear translation, but
alternative explanations for these results have not been
excluded, and other experiments argue against it. Much of
the appeal of nuclear translation is that functional proofreading
of newly made mRNAs in the nucleus would provide an
efficient way to monitor mRNAs for the presence of premature
termination codons, thereby avoiding the synthesis of
deleterious proteins. mRNAs that are still in the nucleusassociated fraction of cells are subject to translational
proofreading resulting in nonsense-mediated mRNA decay and
perhaps nonsense-associated alternate splicing. However,
these mRNAs are likely to be in the perinuclear cytoplasm
rather than within the nucleus. Therefore, in the absence of
additional evidence, we conclude that nuclear translation is
unlikely to occur.
Addresses
Department of Biomolecular Chemistry, University of Wisconsin Medical
School, 1300 University Avenue, Madison, WI 53706, USA
1
e-mail: dahlberg@wisc.edu
Current Opinion in Cell Biology 2004, 16:335–338
This review comes from a themed issue on
Nucleus and gene expression
Edited by Elisa Izaurralde and David Spector
0955-0674/$ – see front matter
ß 2004 Elsevier Ltd. All rights reserved.
DOI 10.1016/j.ceb.2004.03.006
Abbreviations
ESE
exonic splicing enhancer
GFP
green fluorescent protein
NAS
nonsense-associated altered splicing
NMD
nonsense-mediated mRNA decay
PTC
premature termination codon
TCR-b T-cell receptor-b
Introduction
Eukaryotic cells are highly compartmentalized, with
many steps in gene expression being restricted to the
nucleus or to the cytoplasm. This sequestration of functions promotes the controlled and efficient synthesis,
maturation and degradation of macro-molecules and it
also allows monitoring of the integrity of newly made
molecules as they undergo intracellular transport [1].
Most RNAs that function in translation in the cytoplasm
are synthesized and matured in the nucleus, where they
can be destroyed if they are inappropriately processed or
www.sciencedirect.com
incorporated into defective RNPs [2–4]. Also, the integrity of mature ribosomes and tRNAs is monitored during
nuclear export, via their interactions with export receptors
or adaptors [5,6]. Finally, newly processed mRNAs are
functionally proof-read by translation (protein synthesis)
to ensure that they encode full length proteins [7,8].
It is generally assumed that proteins are synthesized only
in the cytoplasm, but several laboratories have challenged
that assumption, proposing that translation can also occur
in the nucleus, at or close to the sites of pre-mRNA
synthesis [9,10–13]. Evidence cited in support of this
assertion includes experiments designed both to look
directly for nuclear translation and to establish communication between translation and nucleus-associated
events. Recently, we critically analyzed data published
in support of, or against, the existence of nuclear translation [14] and pointed out weaknesses in arguments on
both sides of the issue. Here we review recent results
pertaining to this question and conclude that the case
against nuclear translation appears to be getting stronger.
Translation within nuclei?
Clearly, the most direct way to determine if protein
synthesis occurs inside the nucleus would be to ask if
viable, intact isolated nuclei can make proteins. Unfortunately, it is very difficult to isolate nuclei completely
free of contaminating cytoplasmic components, especially
endoplasmic reticulum (ER), which is contiguous with the
nuclear envelope. Because most, if not all, translation
occurs in the cytoplasm, a small amount of contamination
by ER-bound ribosomes could give misleading positive
results. Also, if the nuclei are not intact during isolation
and assay, important translation factors or inhibitors could
leak into or out of them. Thus, multiple controls are needed to show that ‘purified nuclei’ are structurally and
functionally intact and devoid of cytoplasmic ribosomes.
We raised many of these concerns in our previous analysis
of the provocative paper on this subject published by
Iborra et al. [9]. These uncertainties led us to doubt
whether the results presented were relevant to the question of nuclear translation in living cells [14]. At that
time, Nathanson et al. [15] showed that the ability of
isolated nuclei to carry out protein synthesis is reduced
in proportion to the purity of the nuclei. However, this
study did not exclude that the nuclei could have been
damaged during purification, causing inactivation or loss
of an essential translation factor(s), even though they
supported transcription, and thus appeared to be functionally intact. More recently, Herbert and coworkers
[11] observed that in cells containing distorted nucleoli,
Current Opinion in Cell Biology 2004, 16:335–338
336 Nucleus and gene expression
an overexpressed chimeric reporter protein was found
only in the original nucleus of a heterokaryon, when
assayed at short times after cell fusion. Although these
authors argued that nuclear synthesis of the protein was
responsible for this effect, they failed to demonstrate this
claim directly, or to exclude the likely possibility that the
artificial reporter protein exited and entered nuclei very
slowly. In summary, direct evidence for translation within
nuclei is still lacking.
Inherent capacity for nuclear translation
In a slightly less direct manner, arguments for or against
protein synthesis in the nucleus have been made on the
basis of whether components of the translation machinery
(translation factors, tRNAs and ribosomes) can be detected in nuclei. Interpretation of such studies is complex
because the function, rather than just the presence, of
these components in the nucleus needs to be demonstrated. In some cases, the potential for function exists;
for example, tRNAs are synthesized, matured and even
aminoacylated in nuclei [6]. By contrast, newly made
ribosomal subunits in the nucleus are unlikely to function
because they are still immature and do not appear to form
80S ribosomes [5,16,17]. Although it is possible that 80S
ribosomes exist in nuclei, attempts to detect them biochemically or by electron microscopy have not been
successful, and complexes of newly made mRNPs bound
to ribosomes were detected only in the perinuclear
cytoplasm [18].
As we discussed previously [14], the reliability of studies
that search for ribosomes and translation factors near the
site of mRNA formation in the nucleus depends on the
probes used. For instance, Brogna et al. [10] used antibodies of uncharacterized specificity and sensitivity, raising the possibility of false positive (or negative) results.
This problem of specificity and sensitivity of antibodies
can be circumvented by tagging proteins with green
fluorescent protein (GFP), provided that the GFP tag
does not influence the intracellular localization of the
chimeric protein. Quantification of the amounts of several
GFP-tagged translation factors, all of which are subject to
active export from the nucleus, led Bohnsack et al. [19] to
conclude that the nuclear levels of most factors are likely
to be too low to support protein synthesis (see also [20]).
Moreover, Björk et al. [18] reported that the translation
initiation factor eIF4H, which can be detected in the
nucleus, was found associated with newly made Balbiani
ring mRNPs only in the perinuclear cytoplasm, where
loading of ribosomes occurred. On balance, we feel that
the data on the availability of 80S ribosomes and required
translation factors make it unlikely that protein synthesis
can occur in the nucleus, at least under most conditions.
are monitored for the presence of nonsense (translation
stop) codons within their coding regions while the
mRNAs are still in the nucleus-associated fraction of cells
[2,7,8]. Proofreading of newly made mRNAs for premature termination codons (PTCs) relies on translation
because it is sensitive to inhibitors of protein synthesis,
changes in the reading frame or the introduction of a
nonsense-suppressor tRNA. In general, detection of a
PTC results in significant degradation of the mRNA by
nonsense-mediated mRNA decay (NMD). Occasionally,
the presence of a PTC in an mRNA also leads to
nonsense-associated altered splicing (NAS), whereby
alternative mRNAs are produced that lack the exon containing the PTC [21]. Monitoring of mRNAs for PTCs
through translation within the nucleus has the appeal that
the detection and consequences of a PTC would occur in
the same cell compartment [22,23]. (In yeast, NMD
occurs in the cytoplasm so translational proofreading
has no apparent linkage with nuclear events [4].)
If NMD were an intra-nuclear event, as suggested by
Bühler et al. [13], a strong case would be made for some
translation taking place in the nucleus. These authors
showed that NMD occurred even when the rate of mRNA
export was greatly reduced. However, if export was still
faster than NMD, as seems likely, their results would also
be consistent with NMD involving translation immediately after the mRNA exited the nuclear pore [14]. As
noted above, newly exported mRNAs are transiently
associated with perinuclear ribosomes [18], even if the
encoded protein is not destined for the ER lumen
[24,25]. Thus, an mRNA may undergo the ‘pioneer’
round of translation, during which NMD occurs [26], after
it has exited the nucleus but while it is still in the nucleusassociated (perinuclear) cytoplasm. Consequently, the
apparently nuclear location of NMD cannot be taken
as evidence for translation within the nucleus.
PTC-induced alteration of intra-nuclear
events
Detection of a PTC by the translation machinery has
been reported to promote two nuclear events, NAS and
accumulation of the pre-mRNA near the site of transcription. It is unclear whether these two phenomena are
causally related to each other.
Although NAS has been observed for many different premRNAs (reviewed in [21]), in almost all cases, this change
has been shown to be a consequence of mutational
inactivation of an exonic splicing enhancer (ESE), rather
than a result of reading-frame-dependent recognition of a
PTC [27]. However, it can be very difficult to determine
if a PTC mutation elicits NAS solely by inactivation of an
ESE [28].
mRNA proofreading by translation
An often-cited reason for considering the possibility of
nuclear translation is the fact that many spliced mRNAs
Current Opinion in Cell Biology 2004, 16:335–338
The best-documented example of NAS for which inactivation of an ESE appears to have been excluded is the
www.sciencedirect.com
Does protein synthesis occur in the nucleus? Dahlberg and Lund 337
processing of T-cell receptor-b (TCR-b) pre-mRNA
[12]. In this case, the highly polymorphic VDJ exon is
generated by somatic rearrangement of DNA sequences
that encode multiple variable, diversity and junction
(VDJ) sequences, so most of the resulting TCR-b genes
are likely to encode mRNAs with PTCs. However, NAS
and an especially vigorous form of NMD keep the level of
such PTC-containing TCR-b mRNAs low. Interestingly,
Wang et al. [12,29] showed that the reading-framedependent NAS of TCR-b pre-mRNA functions in trans,
because recognition of a PTC generated by normal splicing of one TCR-b pre-mRNA molecule affects the splicing of other molecules. Two models have been proposed
to explain how the translation machinery could communicate with the nuclear pre-mRNA processing apparatus.
In one, translational proofreading of the mRNA occurs at
the site of pre-mRNA processing in the nucleus, and the
detection of a PTC causes, by an unknown mechanism,
the localized excess (or deficit) of factors needed to alter
the splicing pattern of TCR-b pre-mRNAs [22,29]. In the
other model, monitoring of mRNAs for PTCs is entirely
by cytoplasmic ribosomes, and the translation machinery
communicates with nuclear spliceosomes by sequestration of a shuttling, TCR-b-specific splicing factor [14].
A test for the existence of a signaling molecule that
shuttles between ribosomes and spliceosomes is underway
(O Mühlemann, personal communication).
Like TCR-b, the pre-mRNA of Ig-m (which also is the
result of programmed gene rearrangement) has been
reported to accumulate at the site of transcription when
the encoded mRNA contains a PTC [30] and such transcripts may also undergo NAS (discussed in [21,29]). One
would expect that PTC-induced pre-mRNA accumulation, and/or NAS, would be reflected in a change in the
kinetics of splicing of individual introns. However, Lytle
and Steitz [31] found that the presence of several different encoded PTCs had no significant effect on the
rates at which neighboring introns were removed from the
precursors of Ig-m (or DHFR) mRNA [31]. Unexpectedly, these authors also detected considerable variation in
the levels of Ig-m pre-mRNAs amongst different isolates
of the cell line expressing wild-type Ig-m pre-mRNA.
This unexplained variability between control cells raises
questions about the significance of reported increases in
the observed levels of PTC-containing pre-mRNAs [30].
Thus, the evidence for reading-frame-dependent alterations of NAS and/or nuclear pre-mRNA accumulation
should be revisited before these observations can be used
to support proposal that translation occurs within nuclei.
Conclusions
Nuclear translation remains a controversial topic.
Although the papers favoring nuclear translation have
generated a great deal of excitement, they were often
incomplete and required further experimentation. In
spite of considerable effort by several laboratories to
www.sciencedirect.com
demonstrate synthesis of specific proteins at the sites
of transcription of their pre-mRNAs, a ‘smoking gun’
has not yet appeared. Instead, the only recent publications on this subject that we are aware of tend to support
the idea that protein synthesis is restricted to the cytoplasm. Hence, we continue to be skeptical and feel that
there is no reason to embrace the idea that protein
synthesis occurs within cell nuclei.
Update
Recently published data have been interpreted as being
consistent with nuclear translation [32]. However, the
citation of unpublished results of several important
experiments, and a lack of certain controls, allow for other
explanations of the data.
Acknowledgements
Both authors are supported by grant R37-GM-30220 from NIH.
References and recommended reading
Papers of particular interest, published within the annual period of
review, have been highlighted as:
of special interest
of outstanding interest
1.
Maniatis T, Reed R: An extensive network of coupling among
gene expression machines. Nature 2002, 416:499-506.
2.
Stutz F, Izaurralde E: The interplay of nuclear mRNP assembly,
mRNA surveillance and export. Trends Cell Biol 2003,
13:319-327.
3.
Le Hir H, Nott A, Moore MJ: How introns influence and enhance
eukaryotic gene expression. Trends Biochem Sci 2003,
28:215-220.
4.
Vasudevan S, Peltz SW: Nuclear mRNA surveillance.
Curr Opin Cell Biol 2003, 15:332-337.
5.
Johnson AW, Lund E, Dahlberg JE: Nuclear export of ribosomal
subunits. Trends Biochem Sci 2002, 27:580-585.
6.
Lund E, Dahlberg JE: Proofreading and aminoacylation of tRNAs
before export from the nucleus. Science 1998, 282:2082-2085.
7.
Schell T, Kulozik AE, Hentze MW: Integration of splicing,
transport and translation to achieve mRNA quality control by
the nonsense-mediated decay pathway. Genome Biol 2002,
3:1006.
8.
Maquat LE, Carmichael GG: Quality control of mRNA function.
Cell 2001, 104:173-176.
9.
Iborra FJ, Jackson DA, Cook PR: Coupled transcription and
translation within nuclei of mammalian cells. Science 2001,
293:1139-1142.
10. Brogna S, Sato T, Rosbash M: Ribosome components are
associated with sites of transcription. Mol Cell 2002, 10:93-104.
These experiments use in situ hybridization and antibodies of unknown
specificity to show the presence of rRNA and protein components of
translation machinery at sites of active transcription of D. melanogaster
polytene salivary gland chromosomes. Although this reference has been
widely cited for showing that translation and transcription machineries are
near each other in the nucleus, the conditions of analysis and specificities
of the reagents used raise significant questions about the interpretation of
the results (see [14]).
11. Herbert A, Wagner S, Nickerson JA: Induction of protein
translation by ADAR1 within living cell nuclei is not dependent
on RNA editing. Mol Cell 2002, 10:1235-1246.
A study showing that over-expression of various parts of an RNAmodifying enzyme can affect the nuclear localization of a chimeric
protein; the data do not address the assumption stated in the title that
protein translation occurs in living cell nuclei.
Current Opinion in Cell Biology 2004, 16:335–338
338 Nucleus and gene expression
12. Wang J, Hamilton JI, Carter MS, Li S, Wilkinson M: Alternatively
spliced TCR mRNA induced by disruption of reading frame.
Science 2002, 297:108-110.
Experiments demonstrating that NAS works in trans, because a PTC in
one mRNA affects splicing of another, arguing for feedback between
translation and splicing.
13. Bühler M, Wilkinson M, Mühlemann O: Intranuclear degradation
of nonsense-codon-containing mRNA. EMBO Rep 2002,
3:646-651.
An attempt to determine if NMD, and thus translation, occurs in nuclei
before an mRNA is exported to the cytoplasm. Slowing of mRNA export
by treatment of cells with the M protein of VSV did not affect NMD.
However, cytoplasmic monitoring for PTCs in newly exported mRNA
cannot be excluded if NMD is the rate limiting step in the overall process
of mRNA export and proof-reading (see [14]).
14. Dahlberg JE, Lund E, Goodwin EB: Nuclear translation: what is
the evidence? RNA 2003, 9:1-8.
A critical analysis of papers published up to fall, 2002, in which arguments
were made in favor of and against nuclear translation.
15. Nathanson L, Xia T, Deutscher MP: Nuclear protein synthesis:
a re-evaluation. RNA 2002, 9:9-13.
A re-evaluation of experiments [9] that had been interpreted as demonstrating translation within isolated nuclei, showing that the earlier results
could be a consequence of cytoplasmic contamination.
16. Pederson T, Politz JC: The nucleolus and the four
ribonucleoproteins of translation. J Cell Biol 2000,
148:1091-1095.
17. Trotta CR, Lund E, Kahan L, Johnson A, Dahlberg JE: The role of
NMD3 protein in the export of 60S subunits from nuclei of
Xenopus oocytes. EMBO J 2003, 22:2841-2851.
18. Björk P, Bauren G, Gelius B, Wrange Ö, Wieslander L: The
Chironomus tentans translation intitation factore eIF4H is
present in the nucleus but does not bind to mRNA until the
mRNA reaches the cytoplasmic perinuclear region. J Cell Sci
2003, 116:4521-4532.
A study examining whether a translation initiation factor functions when it
is in the nucleus.
19. Bohnsack M, Regener K, Schwappach B, Saffrich R, Paraskeva E,
Hartmann E, Gorlich D: Exp5 exports eEF1a via tRNA from nuclei
and synergises with other transport pathways to confine
translation to the cytoplasm. EMBO J 2002, 21:6205-6215.
Experiments showing that several translation factors are actively exported from nuclei and hence are present in very low amounts within
nuclei (see also [20]).
20. Calado A, Treichel N, Müller E-C, Otto A, Kutay U: Exportin 5
mediated nuclear export of elongation factor 1A and tRNA.
EMBO J 2002, 21:6216-6224.
21. Maquat LE: NASty effects on fibrillin pre-mRNA splicing:
another case of ESE does it, but proposals for translation-
Current Opinion in Cell Biology 2004, 16:335–338
dependent splice site choice live on. Genes Dev 2002,
16:1743-1753.
22. Wilkinson M, Shyu A: RNA surveillance by nuclear scanning?
Nat Cell Biol 2002, 4:E144-E147.
23. Brogna S: Pre-mRNA processing: insights from nonsense.
Curr Biol 2001, 11:R838-R841.
24. Sanchez-Velar N, Udofia EB, Yu Z, Zapp ML: hRIP, a cellular
cofactor for Rev function, promotes release of HIV RNAs from
the perinuclear region. Genes Dev 2004, 18:23-34.
Experiments showing that newly exported mRNAs are associated with
perinuclear structures.
25. Lerner RS, Seiser RM, Zheng T, Lager PJ, Reedy MC, Keene JD,
Nicchitta CV: Partitioning and translation of mRNAs encoding
soluble proteins on membrane-bound ribosomes. RNA 2003,
9:1123-1137.
Experiments showing that a wide variety of mRNAs are transiently
associated with ribosomes that are bound to the endoplasmic reticulum.
26. Ishigaki Y, Li X, Serin G, Maquat LE: Evidence for a pioneer round
of mRNA translation: mRNAs subject to nonsense-mediated
decay in mammalian cells are bound by CBP80 and CBP20.
Cell 2001, 106:607-617.
27. Cartegni L, Chew SL, Krainer AR: Listening to silence and
understanding nonsense: exonic mutations that affect splicing.
Nat Rev Genet 2002, 3:285-298.
28. Caputi M, Kendzior RJ, Beemon KL: A nonsense mutation in the
fibrillin-1 gene of a Marfan syndrome patient induces NMD and
disrupts an exonic splicing enhancer. Genes Dev 2002,
16:1754-1759.
A thorough and careful demonstration of how a mutation that creates a
PTC affects splicing by inactivating an ESE.
29. Wang J, Chang Y, Hamilton JI, Wilkinson MF: Nonsenseassociated altered splicing: a frame-dependent response
distinct from nonsense-mediated decay. Mol Cell 2002,
10:951-957.
30. Mühlemann O, Mock-Casagrande CS, Wang J, Li S, Custodio N,
Carmo-Fonseca M, Wilkinson MF, Moore MJ: Precursor RNAs
harboring nonsense codons accumulate near the site of
transcription. Mol Cell 2001, 8:33-43.
31. Lytle JR, Steitz JA: Premature termination codons do not affect
the rate of splicing of neighboring introns. RNA 2004, in press.
A study showing that the kinetics of intron removal from two pre-mRNAs
are not affected by the presence of a PTC in the product mRNA and that
the levels of ‘control’ pre-mRNAs vary between cell isolates.
32. Iborra FJ, Escargueil AE, Kwek KY, Akoulitchev A, Cook PR:
Molecular cross-talk between the transcription, translation,
and nonsense-mediated decay machineries. J Cell Sci 2004,
117:899-906.
www.sciencedirect.com