ArticleWidespread Accumulation of Ribosome-Associated
Isolated 30 UTRs in Neuronal Cell Populations of the
Aging BrainGraphical AbstractHighlightsd Isolated 30 UTRs accumulate in aging D1 spiny projection
neurons of mouse striatum
30d UTR RNAs occur with oxidative stress and can be induced
by ROS-inducing drugs
0
d 3 UTR RNAs are associated with the aging human brain and
short peptide production
0
d Isolated 3 UTRs may result from impaired ABCE1 and
deficient ribosome recyclingSudmant et al., 2018, Cell Reports 25, 2447–2456
November 27, 2018 ª 2018 The Authors.
https://doi.org/10.1016/j.celrep.2018.10.094Authors
Peter H. Sudmant, Hyeseung Lee,
Daniel Dominguez, Myriam Heiman,
Christopher B. Burge
Correspondence
mheiman@mit.edu (M.H.),
cburge@mit.edu (C.B.B.)
In Brief
Particular brain regions and cell
populations exhibit increased
susceptibility to aging-related stresses.
Sudmant et al. report that fragments of
mRNAs accumulate in the aging brains of
mice and humans. These species are
associated with ribosomes and the
production of small peptides and reflect
regional differences in metabolism and
oxidative stress.
Cell Reports
ArticleWidespread Accumulation of Ribosome-Associated
Isolated 30 UTRs in Neuronal Cell Populations
of the Aging Brain
Peter H. Sudmant,1 Hyeseung Lee,2 Daniel Dominguez,1 Myriam Heiman,2,* and Christopher B. Burge1,3,*
1Department of Biology, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, MA 02139, USA
2Picower Institute for Learning and Memory, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, MA 02139, USA
3Lead Contact
*Correspondence: mheiman@mit.edu (M.H.), cburge@mit.edu (C.B.B.)
https://doi.org/10.1016/j.celrep.2018.10.094SUMMARY
Particular brain regions and cell populations exhibit
increased susceptibility to aging-related stresses.
Here, we describe the age-specific and brain-re-
gion-specific accumulation of ribosome-associated
30 UTR RNAs that lack the 50 UTR and open reading
frame. Our study reveals that this phenomenon im-
pacts hundreds of genes in aged D1 spiny projection
neurons of the mouse striatum and also occurs in the
aging human brain. Isolated 30 UTR accumulation is
tightly correlated with mitochondrial gene expres-
sion and oxidative stress, with full-length mRNA
expression that is reduced but not eliminated, and
with production of short 30 UTR-encoded peptides.
Depletion of the oxidation-sensitive Fe-S cluster
ribosome recycling factor ABCE1 induces the accu-
mulation of 30 UTRs, consistent with amodel in which
ribosome stalling and mRNA cleavage by No-Go
decay yields isolated 30 UTRRNAs protected by ribo-
somes. Isolated 30 UTR accumulation is a hallmark of
brain aging, likely reflecting regional differences in
metabolism and oxidative stress.INTRODUCTION
Aging is characterized by impaired molecular function and the
progressive accumulation of cellular damage (López-Otı́n
et al., 2013). In the human brain, these processes—in combina-
tion with genetic and environmental factors—can result in neuro-
degenerative conditions, such as Alzheimer’s disease and a
general decline in memory and cognitive ability. Neurodegener-
ative diseases tend to impact specific cell populations and
regions of the aging brain—a phenomenon known as selective
neuronal vulnerability (Mattson and Magnus, 2006)—but why
particular cells and brain regions exhibit increased susceptibility
to the stresses of aging and the underlying molecular mecha-
nisms are not well understood.
Recently, in situ hybridization has provided support for the
existence of a class of RNAs in mouse neurons consisting exclu-
sively of the 30 UTRs of mRNAs absent the open reading frameCell Repo
This is an open access article under the CC BY-N(ORF) or 50 UTR (Kocabas et al., 2015). Several isolated 30 UTR
species tested showed differential spatial and temporal distribu-
tion during development, and two were linked with reductions in
the overall level of protein produced from the associated gene.
Several other studies have also reported the phenomenon of iso-
lated 30 UTRs. A study of data from mammalian capped analysis
of gene expression (CAGE) sequencing concluded that these
species were likely the result of post-transcriptional cleavage
(Mercer et al., 2011). Other studies have also concluded that
CAGE sequencing captures some cleavage products, including
a class of 30 UTR species that exhibit both conservation and an
enriched 50 GGG sequence (Carninci et al., 2006; Fejes-Toth
et al., 2009). CAGE-detected cleavage products originating in
30 UTRs also showed substantial overlap (40%) with Ago2- and
Drosha-independent cleavage products identified from tran-
scriptome-wide profiling of endonucleolytic cleavage products
that retain a 50 phosphate (Karginov et al., 2010). Isolated
30 UTRs have also been described in murine immune cells and
human cell lines (Malka et al., 2017). However, the biogenesis
and potential functions of isolated 30 UTRs, and whether they
are of particular importance in the brain are not well understood.
Here, we characterize the distribution of isolated 30 UTRs
throughout the aging mouse and human brain, demonstrating
their widespread accumulation as a function of age. We
observe that these RNAs vary substantially in abundance be-
tween different cell types and regions and are associated
with signatures of oxidative stress. We provide evidence for a
model of isolated 30 UTR biogenesis in which oxidative stress
impairs translation termination, triggering the entry of ribo-
somes into the 30 UTR and endonucleolytic cleavage of mRNAs
near the stop codon by the No-Go decay pathway.RESULTS
Isolated 30 UTRs Accumulate in Specific Aged Neuronal
Cell Types of the Mouse Striatum
The striatum is impacted in both Huntington’s disease and
Parkinson’s disease, both age-dependent neurodegenerative
conditions. To identify cell-type-specific and age-associated
changes in translatingmRNAs thatmight contribute to age-asso-
ciated degeneration of striatal neurons, we profiled ribosome-
associated mRNAs from the two major neuronal cell subtypes
of this region: spiny projection neurons (SPNs) of the direct andrts 25, 2447–2456, November 27, 2018 ª 2018 The Authors. 2447
C-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
A AAAAA D
AAA
A
A
AAAAA
striatum AAAAA
age
PN42 2yr Mixed SPNs, SPN specific GFP TSS termination Poly-A 
aging mouse timecourse interneurons, and glia population tagged polysomes codon (tc)
B Aged D1 SPNs C EPpia Rnf5 Cdpf1
1.00
15 9
10
10 6 0.75
5
5 3
0.50
0 0 0 D1 D2
15 9 0.25 PN42
10
10 6 2yr
5 0.005 3
0 5 10
0 0 0
15 R9 tc
10 F
10 6
1000
5 5 3
800
0 0 0 600
15 9 400
10
10 6 200
5 5 3 0
x= 3 4 5 6
0 0 0
Rtc x
0 200 400 600 800 0 300 600 900 0 250 500 750 1000
Figure 1. Isolated 30 UTRs Accumulate in Aged D1 SPNs of the Mouse Striatum, but Not D2 SPNs or Young SPNs
(A) Overview of the TRAP protocol used to isolate cell-type-specific mRNA from D1 and D2 SPNs of the mouse striatum in young and aged mice.
(B) Heatmap of normalized read coverage over the 250most 30 UTR-enriched genes (see below) in 2-year-old D1 SPNs centered on the stop codon and extending
50 and 30 to the TSS and poly(A) site, respectively.
(C) Examples of the log10 coverage of four biological replicates per condition stacked on top of each other across a control gene, Ppia, and two genes exhibiting
isolated 30 UTRs exclusively in aged D1 SPNs. Dotted line indicates the stop codon.
(D) Schematic of the Rtc metric.
(E) The cumulative distribution of Rtc values of genes among the mouse SPN samples reveals a long tail for aged D1 SPNs.
(F) Counts of the number of genes at various Rtc cutoffs in aged D1 SPNs.
reads/ntlo (glo(cgvg)10 stacked)
coverage
2YR PN42
D2 D1 D2 D1
cumulative fraction
No. of genes 
(aged D1 SPNs)
Drd1a.PN42 Drd2.PN42 Drd1a.2yr Drd2.2yr
Drd1a.PN42 Drd2.PN42 Drd1a.2yr Drd2.2yrindirect pathways—referred to hereafter as D1 and D2 SPNs,
respectively. Cell-type-specific translating mRNAswere isolated
in young and aged mice by using translating ribosome affinity
purification (TRAP), which makes use of transgenic mouse lines
expressing an EGFP-tagged L10a ribosomal protein gene driven
by a cell-type-specific promoter (Heiman et al., 2008). D1 and D2
SPNs express almost exclusively Drd1a or Drd2 (dopamine re-
ceptors 1 and 2), respectively, the regulatory elements of which
were used to drive expression of the tagged L10a protein. This
system thus allows for cell-type-specific isolation of ribosome-
associated mRNA from each of these cell types, avoiding
contamination from non-neuronal cell types (Figure 1A).
TRAP RNA was isolated in biological quadruplicate from
42-day-old (PN42) and 2-year-old D1- and D2-tagged mice,
and ribosome-associated (non-poly(A)-selected, rRNAdepleted)
RNA was sequenced. This sequencing procedure captures all
ribosome-associated RNAs, including those without a poly(A)
tail in a highly specific fashion (Figure S1A). In addition to gene
expression changes across thousands of genes (see below), a
meta-gene analysis of sequencing coverage profiles aligned to
the stop codon indicated a dramatic increase in sequence reads
at and beyond the stop codon in aged D1 SPNs (Figures 1B and
S1B). This pattern was not observed in aged D2 SPNs or in sam-
ples from youngmice (Figure 1C). An inspection of read coverage
profiles identified a subset of genes in aged D1 SPNs that had2448 Cell Reports 25, 2447–2456, November 27, 2018read coverage exclusively or mostly restricted to the 30 UTR.
This pattern was highly consistent across biological replicates
(Figure S1C). We confirmed a several-fold increased abundance
of 30 UTRs relative to coding regions in aged D1 SPNs for several
of these genes by qPCR analysis (Figure S1D), suggesting that
full-length (or ORF-only) mRNAs are reduced but not absent
for this set of genes. The signatures associated with isolated
30 UTRs observed here in aged D1 SPNs are of much greater
magnitude and extend across far more genes than has been
previously reported (Kocabas et al., 2015; Malka et al., 2017).
To quantify the extent of 30 UTR enrichment of a gene in a data-
set, we developed a simple metric that we call the ‘‘termination
codon ratio,’’ Rtc, defined as the log2 of the ratio of read density
in the 30 UTR of a gene to read density in the body of the gene
(including coding region and 50 UTR, STAR Methods), restricted
to constitutive portions of the gene (Figure 1D). Rtc is analogous
to measures used previously to quantify ribosome occupancy in
30 UTRs from ribosome footprint data (Mills et al., 2016), but
when applied to TRAP or RNA-seq data, Rtc measures the
positional enrichment of transcript fragments rather than ribo-
some abundance. The distribution of Rtc values was approxi-
mately symmetrical, with a mean slightly above 0 across young
D1 SPNs (mean Rtc = 0.44) and aged D2 SPNs (mean
Rtc = 0.38), but was substantially skewed toward larger values
in aged D1 SPNs (mean Rtc = 1.34, p < 2.2e16, Wilcoxon
A B
C
D E
Figure 2. Increased Levels of Oxidative Stress Accompany Isolated
30 UTR Accumulation in Aging D1 SPNs
(A) Venn diagram depicting the number of genes differentially expressed be-
tween D1 and D2 SPNs at each time point and the log-fold intra-cell-type gene
expression changes from PN42 to 2 years.
(B) The number of core components of oxidative phosphorylation differentially
expressed between young and old D1 and D2 SPNs. p values indicate
enrichment over background. ***p < 0.001, hypergeometric test.
(C) Representative images of DAPI, lipofuscin, GFP, and merged channels for
9-week and 19-month-old CP73 mice expressing EGFP-L10a in D1 SPNs.
(D) The proportion of lipofuscin-positive cells in GFP ± cells in CP73 and
CP101 mouse lines at 9 weeks. Each point represents a slide and the
x-coordinate was jittered to improve readability. ***p < 0.001, chi-squared test.
(E) The cumulative distribution of the proportional area of each cell positive for
lipofuscin accumulation in 9-week and 19-month-old mice.rank-sum test; Figure 1E). Among aged D1 SPNs, 404 genes ex-
hibited an Rtc R5, indicating at least 2
5 = 32-fold enrichment of
reads in 30 UTR relative to gene body, versus 66 and 52 genes
observed at this cutoff in young D1 SPNs and aged D2 SPNs,
respectively (Figure 1F). Genes with increased Rtc represented
a variety of functions and pathways (none strongly enriched)
and were generally shorter, with a higher C+G content and an
increased potential for 30 UTR RNA secondary than expressed
genes overall (Figures S1E and S1F).Although the presence of 30 UTR RNAs lacking the coding
region has been observed previously in some instances (Carninci
et al., 2006; Kocabas et al., 2015; Malka et al., 2017; Mercer
et al., 2011), the mechanism of their biogenesis remains elusive.
Analyzing total RNA sequencing data from several mouse tis-
sues and the mouse NIH 3T3 cell line, we identified several
genes, including Frat2, Hnrnpa0, and Sox12, that had read den-
sities at least several-fold higher in the 30 UTR than in the coding
region in most or all of these samples (Figures S1G and S1H).
Thus, the phenomenon of 30 UTR-enriched genes observed so
widely in aged D1 SPNs extends to some genes in other mouse
cells and tissues, as observed previously (Kocabas et al., 2015;
Malka et al., 2017).
Aged D1 SPNs Exhibit Increased Oxidative Damage
Compared to D2 and Young SPNs
Because gene expression signatures can provide information
about cellular environmental conditions, we examined gene
expression profiles of young and aged mouse SPNs for clues to
the conditions that give rise to 30 UTR enrichment. Young D1
and D2 SPNs exhibited cell-type-specific expression signatures
typical of those that have been observed previously bymicroarray
and single-cell sequencing studies (Gokce et al., 2016; Heiman
et al., 2008). However, 2-year-old mice exhibited three times as
many differentially expressed genes as PN42 mice (2526 genes
compared to 835; Figure 2A). Changes in expression levels
between young and aged D1 SPNs, rather than changes in D2
SPNs, explained most of these differences, and a pathway anal-
ysisofgenesdifferentiallyexpressedbetweenagedSPNsubtypes
showed an enrichment in nuclear-encoded oxidative phosphory-
lation (OXPHOS) components (Figures 2B and S2A–S2C). We
also observed increased expression of the primary mitochondrial
superoxide dismutase Sod2, which plays crucial roles in elimi-
nating reactive oxygen species (ROS) and sensing oxidative phos-
phorylation-associated distress (Zou et al., 2017) and changes
in other reactive oxygen response genes in aging D1 SPNs (Fig-
ure S2D). These observations raised the possibility that D1 SPNs
might be exposed to higher levels of oxygen free radicals.
To ask whether D1 SPNs are exposed to higher levels of oxida-
tive stress, we quantified lipofuscin aggregates, a byproduct
of oxidatively damaged proteins, lipids, and mitochondria, in
sectioned striata of both young and agedmouse lines expressing
GFP in either a D1-specific (CP73) or D2-specific (CP101) manner
(Figure 2C). Some 95% of the neuronal cells of the striatum are
SPNs,allowingus todirectly compare theaccumulationof lipofus-
cin in GFP-positive versus GFP-negative neurons in each of these
mouse lines. Both aged D1 and aged D2 SPNs exhibited a signif-
icantly higher average per cell lipofuscin fraction than youngSPNs
(p < 1e10, chi-square test), consistent with previous observa-
tions of lipofuscin aggregate accumulation in aging neurons
(Nandy, 1971).However, at 9weeks, 47.5%ofD1SPNswerepos-
itive for lipofuscin speckles compared to 37.8% of D2 SPNs (Fig-
ure 2D; p < 1e10, chi-square test). Furthermore, the mean per
cell lipofuscin fraction was also significantly higher in aged D1
SPNs compared to D2 SPNs (Figure 2E; p = 0.002, Wilcoxon
rank-sum test). These observations support the idea that wild-
type D1 SPNs are subject to greater exposure to oxygen free
radical species during aging than D2 SPNs.Cell Reports 25, 2447–2456, November 27, 2018 2449
A Model of Isolated 30 UTR Formation from Impaired
ABCE1 Activity and Cleavage by the No-Go Decay
Pathway
The association of ribosomes with isolated 30 UTRs, implied
by their isolation via TRAP, and the location of the increase in
30 UTR enrichment close to the stop codon are consistent with
a cytoplasmic, translation-linked mechanism for biogenesis of
isolated 30 UTRs.We considered a number of translation-associ-
ated mechanisms involving the 30 UTR as possible sources of
isolated 30 UTRs. For example, nonsense-mediated mRNA
decay (NMD) targets mRNAswith premature termination codons
and/or long 30 UTRs for degradation (Garneau et al., 2007). How-
ever, we observed no strong relationship between 30 UTR length
and 30 UTR enrichment (Figure S1F), and NMD is associated
with prematurely terminating ribosomes rather than 30 UTR-
associated ribosomes. Alternatively, translational read-through
(i.e., translation extending through the stop codon) would yield
30 UTR-associated ribosomes. However, read-through by itself
does not produce mRNA fragments. Recent studies in yeast
andmammalian hematopoietic cells have observed the accumu-
lation of ribosomes in 30 UTRs following depletion of the yeast
ribosome recycling factor Rli1 (Young et al., 2015) and of its hu-
man homolog, ABCE1 (Mills et al., 2016). Depletion of ABCE1
leads to the increased accumulation of ribosomes on the stop
codon and subsequent reinitiation of translation in the 30 UTRs
of a subset of genes (Guydosh and Green, 2014; Young et al.,
2015); the extent of this phenomenon varies widely between
genes for reasons that are not understood. The failure of
ABCE1 to efficiently recycle ribosomes on particular mRNAs
has been associated with increased rates of both read-through
and reinitiation, although mechanistic details are lacking (Guy-
dosh and Green, 2014; Mills et al., 2016; Young et al., 2015).
A model in which the formation of 30 UTR RNA fragments re-
sults from impaired ABCE1 activity is also consistent with the
observed association between oxidative stress and 30 UTR
enrichment. ABCE1 contains two iron-sulfur clusters that are
essential for its activity and are acutely sensitive to oxygen free
radicals (Alhebshi et al., 2012). Treatment with ROS-inducing
compounds, such as paraquat (PQ), impairs ABCE1 iron-sulfur
cofactor uptake, reducing its activity (Alhebshi et al., 2012). In
yeast, age-induced mitochondrial dysfunction leads to impaired
assembly of iron-sulfur clusters, which are synthesized in the
mitochondrial matrix, and targeted mRNA decay of numerous
iron-dependent genes, including Rli1 (Puig et al., 2005; Veatch
et al., 2009). The induction of oxidative stress with hydrogen
peroxide in yeast also increases the proportion of 30 UTR ribo-
somes, consistent with defects in ribosome recycling (Gerash-
chenko and Lobanov, 2012).
Transcripts that fail to undergo translation termination and
have stalled ribosomes are normally degraded through the
PELO/HBS1-dependent No-Go decay (NGD) pathway, which
induces endonucleolytic cleavage of the message just upstream
of the stalled ribosome (Doma and Parker, 2006; Tsuboi et al.,
2012). Cleavage of the mRNA yields two fragments, with the
50 fragment degraded 30-to-50 by the exosome and the 30 frag-
ment normally degraded 50-to-30 by XRN1 (Tsuboi et al., 2012).
This process has also been implicated inmRNAdegradation dur-
ing NMD (Arribere and Fire, 2018). A similar process of ribosome-2450 Cell Reports 25, 2447–2456, November 27, 2018associated endonucleolytic cleavage just upstreamof ribosomes
termed ‘‘ribothrypsis’’ has recently been reported and could be
considered a generalized form of NGD (Ibrahim et al., 2018).
We hypothesized that the consistent failure of ABCE1 to
recycle a ribosome from the stop codon of an mRNA and from
its 30 NGD cleavage product could produce a relatively stable
30 UTR RNA species. Such an RNA would be protected from
exonucleases at its 50 end by a persistently stalled ribosome
(and perhaps secondarily stalled ribosomes upstream) and at
its 30 end by the poly(A) tail (Figure 3A). Inhibition of ABCE1 is
known to increase the rates of read-through and reinitiation
(Guydosh and Green, 2014; Mills et al., 2016; Young et al.,
2015), potentially giving rise to secondary RNA species. Read-
through or reinitiation involves ribosome movement into the
30 UTR, potentially yielding partial 30 UTR fragments protected
at their 50 ends by secondarily stalled ribosomes (Figure 3A).
Depletion of ABCE1 Induces Cleavage-Associated
Fragments and 30 UTR Enrichment
To directly test our hypothesis that reduced ABCE1 activity can
trigger 30 UTR enrichment, we used small interfering RNAs
(siRNAs) to knock down Abce1 in cultured NIH 3T3 cells (Figures
3B and S3A) and performed qRT-PCR with primers to the
Nanos1 gene, which showed robustly increased 30 UTR enrich-
ment in aged D1 SPNs. This knockdown resulted in a 50% in-
crease in 30 UTR enrichment of Nanos1 after 3 days, assessed
using primer pairs in the 30 UTR versus pairs in the ORF. To
test whether oxidative stress could inhibit ABCE1 and subse-
quently induce isolated 30 UTRs, we treated 3T3 cells with the
drug PQ, a potent inducer of ROS. Over a 5-day treatment
time course with 100 mM PQ, ABCE1 protein levels decreased
by 4-fold (Figure S3B), possibly by limiting or chelating iron
and inhibiting the assembly or stability of ABCE1 (Brumaghim
et al., 2003), demonstrating impacts of oxidative stress on the
ABCE1 protein. Furthermore, PQ-induced ABCE1 depletion
was accompanied by the increased 30 UTR enrichment of
Nanos1 (measured after 72 hr; Figure 3B). Thus, both direct
depletion and indirect oxidative stress-induced inhibition of
ABCE1 activity can induce 30 UTR accumulation.
To test the above model, we performed ribosome profiling in
Abce1 knockdown cells (‘‘siA’’) and under induced oxidative
stress by using PQ as described above. Results were compared
to control siRNA (‘‘siC’’). Both PQ and siA treatments reduced
ABCE1 levels by >90% (Figures S3A and S3B). Ribosome-pro-
tected RNAs were isolated from these samples and sequenced
using standard techniques targeting ribosome-protected RNA
fragments 15–32 nt long (STAR Methods).
To identify bulk changes in ribosome positioning in response
to our treatment conditions, we generated metagene plots of
the density of ribosome footprint 50 ends from fragments 27–31
nt long, corresponding to the width of a ribosome. These plots
showed the expected 3-nt periodicity over transcripts, and
metagene plots centered on the stop codon showed a higher
density of stop codon stalled ribosomes after knockdown of
Abce1 (Figure S3C), as reported previously (Mills et al., 2016;
Young et al., 2015). We also observed a substantial increase in
the density of 30 UTR ribosomes after Abce1 knockdown and
PQ treatment (Figure 3C).
A B C
D E
F G
Figure 3. Oxidative Stress and ABCE1 Depletion Induce 30 UTR Enrichment and a Model for Isolated 30 UTR Accumulation
(A) A model of isolated 30 UTR RNA species formation. An excess of age- or drug-induced ROS impairs the activity of the iron-sulfur cluster protein ribosome
recycling factor ABCE1. ABCE1 deficiency results in an increase in ribosomes stalled on the stop codon and reinitiating in the 30 UTR. Transcripts with stalled
ribosomes are endonucleolytically cleaved upstream of stalled ribosomes by the No-Go decay pathway. The 50 cleavage product is degraded by the 30-to-50
exosome, but the 30 fragment is protected from Xrn1 50-to-30 digestion by stalled ribosome(s) at the stop codon.
(B) qPCR analysis of 30 UTR abundance normalized to primer pairs in the coding region for i) cells treated for 3 days with siRNAs targeted to Abce1 compared to
cells transfected with control siRNAs and ii) cells treated for 3 days with 100 mM paraquat compared to untreated cells. Error bars indicate 95% confidence
interval, * indicates significance by bootstrap (p < 0.01).
(C)Meta plot of the density (in average reads permillion, RPM) of 50 ends of 27- to 31-nt-long ribosome protected fragments (RPF) plotted as a function of distance
from the stop codon. p values represent Wilcoxon Rank Sum test for the first 250 nt after the stop codon.
(D) Mean RPM +SEM in the 30 UTRs of genes, binned by Rtc in aged D1 SPNs. p values represent theWilcoxon Rank Sum test between the highest (7
th) bin of the
control condition (siC) and the highest bin of each of the test conditions.
(E) Ribosome footprint density over the first 200 nt of the 30 UTR from genes in the top and bottom 15% of aged D1 SPN Rtc values.
(F) Schematic of cleavage upstream of a stalled ribosome and subsequent protection of a short cleavage fragment by an upstream translating ribosome and the
proportion of short cleavage fragments sequenced from each experimental condition.
(G) Smoothed meta plot of the density (in RPM) of short (14–16 nt) ribosome-protected fragments in relation to the stop codon for different experimental
conditions. Insets are mean + SEM of RPM in the 200-nt upstream and downstream of the stop codon.Our model of isolated 30 UTR formation predicts that siABCE1
(small interfering RNA to ABCE1) or PQ treatment should induce
30 UTR ribosomes specifically in genes susceptible to 30 UTR
enrichment compared to other genes. To test this hypothesis,
we compared the density of 30 UTR ribosomes in genes with
high Rtc values in aged D1 SPNs to genes with low Rtc values
in aged D1 SPNs (top and bottom 15%). The average densityof 30 UTR ribosomes induced by PQ or siABCE1 was 3- to
4-fold higher in high Rtc genes compared to low Rtc genes and
was shifted significantly higher relative to control treatments
(Figures 3D and 3E). These observations support ourmodel, indi-
cating that 30 UTR ribosomes are induced under conditions that
trigger 30 UTR accumulation specifically in those genes, which
exhibit aging-associated increases in Rtc values.Cell Reports 25, 2447–2456, November 27, 2018 2451
The bulk of ribosome-protected fragments recovered from our
ribosomeprofilingassaywere27–31nt long, corresponding to the
region protected by an 80S ribosome. However, a small propor-
tion (0.5%–2.5%) of recovered fragments corresponded to a
second peak at 15 nt (Figure S3D). These short fragments
have been shown to result from cleavage upstream of a stalled
ribosome and subsequent protection by an upstream translating
ribosome, which stalls when the A site of the ribosome reaches
the cleavage site (Figure 3F; Guydosh and Green, 2014; Young
et al., 2015). This stalled ribosomecan in turn inducemRNAcleav-
age upstream, resulting in a chain reaction for hundreds of nucle-
otidesupstreamof the initial cleavageevent and theenrichment of
short ribosome-protected fragments throughout this region.
We quantified short cleavage fragments as a proportion of the
total number of recovered ribosome-protected fragments (Fig-
ure 3F). ABCE1 knockdown and PQ treatments increased the
proportion of cleavage fragments by 2- to 3-fold. A metagene
plot of these short cleavage fragments centered at the stop
codon exhibited a peak just upstream of the stop codon and
extending 200-nt upstream (Figure 3G). This pattern was also
observed in a reanalysis of the shABCE1 (short hairpin RNA to
ABCE1) knockdown in K562 cells (Figures S3E and S3F; Mills
et al., 2016). Short cleavage fragments were enriched to the
largest extent just upstream of the stop codon, whereas PQ
and ABCE1 knockdown increased the density of short cleavage
fragments both upstream and downstream of the stop codon
(Figure 3G, inset), consistent with the variability observed in
the precise boundaries of 30 UTR enrichment. Similar to full-
length ribosome-protected fragments, short cleavage fragments
were more enriched in response to PQ or ABCE1 knockdown in
the 30 UTRs of high Rtc genes (Figure S3G). These results sug-
gest that endonucleolytic transcript cleavage occurs frequently
near stop codons and is increased under conditions of oxidative
stress or direct depletion of ABCE1 for a subset of mRNAs.
Endonucleolytically cleaved mRNAs are susceptible to degra-
dation by 50 terminal exonuclease (TEX) treatment because they
are not protected at their 50 ends by a cap. We isolated total RNA
from the above cellular conditions and sequenced this material
before and after TEX treatment. In untreated samples, we iden-
tified a subset of candidate induced 30 UTR geneswith increased
Rtc relative to the control in both the Abce1 knockdown and PQ
conditions (Figure S3H). In the TEX-treated cells, however, no
difference in Rtc was observed among control, PQ, and Abce1
knockdown treatments, consistent with high Rtc values resulting
from endonucleolytic cleavage.
Together, these observations provide support for our model in
which age-associated oxidative stress and the subsequent
impairment of the activity of the ribosome recycling factor
ABCE1 triggers 30 UTR ribosomes, mRNA cleavage in the vicinity
of the stop codon, and, ultimately, the accumulation of isolated 30
UTRs for many genes in the aging brain. However, the observed
patterns are complex and other processesmay contribute aswell.
Isolated 30 UTRs Accumulate in the Aging Human Brain
and Vary Between Brain Regions
To determine whether isolated 30 UTRs accumulate in the aging
human brain, we analyzed RNA-seq data from various human
cells and tissues from individuals of varying ages. An analysis2452 Cell Reports 25, 2447–2456, November 27, 2018of the total RNA sequencing data from immunopanned human
astrocytes derived from 16 individuals (Zhang et al., 2016) iden-
tified dozens of genes with read coverages several-fold higher
than their 30 UTRs, with a pronounced increase in read density
beginning at or just beyond the stop codon, as illustrated for
SOX9 (Figure 4A). Furthermore, the extent of 30 UTR enrichment
in SOX9 increased with age (Figure 4A, inset; p = 0.00016,
F-test). More broadly, an analysis of 1,380 poly(A)-selected
RNA-seq samples representing 30 different tissues collected
as part of the Genome-Tissue Expression (GTEx) project (GTEx
Consortium, 2015) identified brain region- and age-specific pat-
terns of 30 UTR enrichment. The Rtc values among brain-derived
tissues were positively associated with age, with a significantly
increased slope compared to non-brain tissues (Figures 4B,
S4A, and S4B; p < 2.2e16, F-test). Under our model of isolated
30 UTR biogenesis, reduced ABCE1 activity should be associ-
ated with higher levels of 30 UTR enrichment. Consistent with
this prediction, among the 320 GTEx human neuronal tissues,
ABCE1 expression was significantly lower in samples with higher
mean Rtc (Figure 4C, p = 1.7e25, F-test). ABCE1 gene expres-
sion levels also declined significantly as a function of sample age
(Figure S4C; p = 2.9e6, F-test) but correlated more strongly
with sample Rtc than with age.
Among the 12 different regions of the brain analyzed by the
GTEx consortium, the slope of Rtc versus age was generally pos-
itive but exhibited significant variation (p < 2.2e16, ANOVA), with
only the cerebellumexhibiting a negative correlationwith age (Fig-
ure 4D). These findings suggest that, although 30 UTR enrichment
is associated with increased age, this phenomenon is muchmore
prevalent in neuronal than non-neuronal tissues, with different
neuronal cell types exhibiting this phenomenon to different ex-
tents. In the aging brain, signatures of oxidative damage accumu-
late in more metabolically active brain regions (Corral-Debrinski
et al., 1992). In particular, regional levels of aerobic glycolysis
correlate spatially with amyloid-b deposition and are associated
with increased levels of oxidative stress and susceptibility to
neurodegenerative disease (Vlassenko et al., 2010). We noted
that the regions of the brainwith the highest and lowestbasal rates
ofaerobicglycolysis (Goyal etal., 2017)—the frontal cortexand the
cerebellum, respectively—exhibited the highest and lowest levels
of age-associated 30 UTR accumulation, respectively. These ob-
servations suggest that metabolic differences may contribute to
brain region-specific differences in 30 UTR accumulation.
Neuronal Isolated 30 UTR Accumulation Correlates with
Mitochondrial Gene Expression
The aging human brain is known to experience sustained expo-
sure to ROS (Mattson and Magnus, 2006). We sought to deter-
minewhether signaturesof increasedoxidative stresswereasso-
ciated with increased 30 UTR enrichment. To identify genes
associated with isolated 30 UTRs, we calculated the mean Rtc
value across genes in eachGTExbrain sample as a summary sta-
tistic. This vector of mean Rtc values was then correlated to the
vector of gene expression values in matched samples, for every
gene. The resulting correlations and p values were subsequently
ranked to identify genes of interest (Figure 4E). Strikingly, 45 of
the 100 genes most positively correlated with mean per sample
Rtc across GTEx samples were core mitochondrial genes or
A B C
D E
Figure 4. 30 UTR-Enriched Genes Accumulate in the Aging Human Brain and Are Associated with Translation of Short Peptides
(A) Stacked plot of the log10 coverage of Sox9 in human astrocytes colored by age. Inset is the Rtc value plotted as a function of age (p value by F-test).
(B) The relationship between age and the mean Rtc in 1380 GTEx samples colored by brain and non-brain tissues and the linear fit between age and Rtc across all
genes for brain and non-brain tissues. For each individual sample, the mean Rtc and confidence intervals about the mean are plotted versus age.
(C) Themean Rtc in neuronal GTEx samples plotted against expression of ABCE1 and boxplots of the distribution of ABCE1 gene expression in different neuronal
tissues. Samples with decreased ABCE1 levels exhibit significantly higher average Rtc (p value by F-test).
(D) The relationship between age and Rtc separated by brain region.
(E) The top 100 genes most positively correlated with increased Rtc in neuronal GTEx samples by rank and p value. Insets are a donut plot demonstrating the
distribution of functional classifications of genes and a histogram of the p value distribution of positive gene correlations.oxidative stress response genes, including 16 mitochondrial-en-
coded transcripts. Applying the same approach toRNA-seq data
from cultured mouse cortical neurons treated with a panel of
drugs (Figure S4D), we again observed that mitochondrial genes
were among themost strongly correlated with increasedRtc (Fig-
ureS4E). In both of these analyses, althoughmitochondrial genes
encoded on both the nuclear and mitochondrial genomes were
robustly enriched, mitochondrial-encoded transcripts ranked
most highly. These results confirm an association between 30
UTR enrichment and upregulation of mitochondrial genes and
oxidative stress response genes in both humans and mice.
Evidence that 30 UTR-Encoded Peptides Originate from
Isolated 30 UTR Genes
In both human and mouse, some 30 UTR sequences accumulate
more than others (Figure 1F and 4B). By examining the 30 UTR
sequences of high Rtc genes, we observed that genes with
higher Rtc were most likely to have TGA stop codons and leastlikely to have TAA (Figures S5A and S5B; p < 1e10, D1 SPNs;
p = 0.0043, GTEx, Cochrane-Armitage Test). A stop codon
sequence can influence the fidelity of termination (Dabrowski
et al., 2015) and may contribute to gene-specific differences in
isolated 30 UTR accumulation. We also observed that themedian
distance from the canonical stop codon to the first 30 UTR stop
codon (in any frame) was 2-fold longer for genes with higher
30 UTR enrichment in both humans and mice (but still much
shorter than the main ORF) (Figure 5A). This observation raised
the possibility that 30 UTR ribosomes may be actively translating
in some cases. In yeast, 30 UTR-encoded peptides are produced
when the ABCE1 homolog Rli1 is inhibited (Young et al., 2015).
To explore the possibility that peptides are produced from
ribosomes associated with isolated 30 UTRs in mammals, we
analyzed tandem mass-spectrometry (MS/MS) data derived
from post-mortem human brain surveyed across 7 different tis-
sues (Carlyle et al., 2017). Spectra were mapped to both the Uni-
Prot protein database aswell as a custom database of translatedCell Reports 25, 2447–2456, November 27, 2018 2453
A B D
C
Figure 5. Evidence of 30 UTR-Encoded Peptides Originating from Isolated 30 UTR Genes
(A) Box plots of the distribution of distances from the stop codon to the next 30 UTR stop codon (in any frame) plotted as a function of Rtc in 7 equal sized bins.
(B) Example of translated 30 UTR ORFs from the 30 UTRs of PQLC1 and (c) SLC2A6. Red vertical dashes indicate stop codons in frames relative to the canonical
stop codon. Red text indicates the peptide spectra that mapped to the 30 ORF, which is shown in orange.
(C) Example of translated 30 UTRORFs from the 30 UTRs of SLC2A6. Red vertical dashes indicate stop codons in frames relative to the canonical stop codon. Red
text indicates the peptide spectra that mapped to the 30 ORF, which is shown in orange.
(D) Box plots of the mean Rtc for all genes compared to those with 3
0 UTR peptides (p = 4.9e7, t test).30 UTR ORFs (STAR Methods). As a positive control for our
computational pipeline, we detected an average of 9,648 high-
confidence annotatedproteins per sample, consistentwith previ-
ous analyses. We also detected 470 high-confidence 30 UTR
ORF-derived peptides (at a false discovery rate of 0.01), which
were on average 52 amino acids long and lacked recognizable
protein domains (e.g., see Figures 5B, 5C, and S5C). Genes
with associated 30 UTR peptides had significantly higher Rtc
values compared to similarly expressed genes whose annotated
proteins were detected by mass spectrometry (Figure 5D,
p = 4.9e7, t test), suggesting that these peptides may be pro-
duced from isolated 30 UTRs or that the production of isolated
30 UTRs and 30 UTR-encoded peptides involves related mecha-
nisms. These peptides were derived from all three ORFs relative
to the upstream canonical ORF in similar proportions (Fig-
ure S5D). Thus, the active translation of isolated 30 UTRs may
be a biological consequence of increased oxidative stress, but
whether the resulting peptides have function is not clear.
DISCUSSION
Here, we have documented the widespread accumulation of iso-
lated 30 UTRRNAspecies, observed several times previously, and
we demonstrate an association with age and present a model for
their biogenesis (Figure 3A). Althoughothermechanismsmay also
contribute to isolated 30 formation, our model proposes that a
consequence of age-associated oxidative stress is the impair-
ment of the ribosome-recycling factor ABCE1 and endonucleo-
lytic cleavage near stop codon-arrested ribosomes. Four main2454 Cell Reports 25, 2447–2456, November 27, 2018lines of evidence support our proposedmechanism. First, isolated
30 UTRs are consistently associated with, and can be directly
inducedby, oxidative stress. Second, impairment of the ribosome
recycling factor ABCE1, which occurs under conditions of oxida-
tive stress, induces 30 UTR accumulation. Third, inhibition of
ABCE1 either directly or by oxidative stress increases the abun-
dance of 30 UTR ribosomes specifically in those genes that exhibit
isolated 30 UTR production. Finally, inhibition of ABCE1 directly or
by oxidative stress increases the proportion of short cleavage-
associated ribosome-protected fragments at and around the
stop codon, providing direct evidence that isolated 30 UTRs are
produced by cleavage of larger RNAs. Thus, the transcriptome
may be reshaped by No-Go decay under stress.
Isolated 30 UTRs appear to be a cell-type- and region-specific
biomarker of neuronal aging and oxidative stress. Although
specific functions are not known, the broad scale of this phenom-
enon is likely to contribute to one or more biologically important
processes. Because the production of isolated 30 UTRs is pre-
sumably accompanied by degradation of the coding portion of
the message, downregulation of the expression of the associated
proteins is likely, as has been observed previously (Kocabas et al.,
2015). This downregulation likely has a variety of biological conse-
quences, independent of any role of the isolated 30 UTR species.
Furthermore, the accumulation of 30 UTRs could itself have signif-
icant impact.mRNA30 UTRs playmany roles in the cell, regulating
mRNA localization, stability, and translation via regulatory ele-
ments (Kuersten and Goodwin, 2003). The presence of excess
30 UTR sequences in the cell could thus sequester RNA-binding
proteins or microRNAs, analogous to some circular RNAs or to
CUG repeat RNAs in myotonic dystrophy (Goodwin et al., 2015;
Hansen et al., 2013). The 30 UTRmay also have a scaffolding func-
tion, directing protein-protein interactions (Berkovits and Mayr,
2015), and free 30 UTRs might perturb these functions.
The physical association between isolated 30 UTRs and ribo-
somes suggests additional potential functions. For example, iso-
lated 30 UTRs may globally reduce translation efficiency by
sequestering large numbers of ribosomes. Additionally, 30 UTR-
derived peptides might have cellular consequences. A recent
study identified short peptides generated from upstream ORFs
(uORFs) during the integrated stress response thatwerepredicted
to have high affinity for major histocompatibility class I antigens of
the adaptive immune system (Starck et al., 2016). Peptides gener-
ated from 30 UTRs in the aging brain could similarly elicit immune
responses, contributing to age-associated inflammation, or could
contribute to the formation of protein aggregates or other proteo-
toxic stresses commonly observed in aging brains.
The relationship between aging and isolated 30 UTRs has in-
clined us toward potential detrimental consequences of 30 UTR
accumulation. However, it is also possible that isolated 30 UTRs
contribute to adaptive cellular responses. In Salmonella, endo-
nucleolytic cleavage of the cpxP 30 UTR produces a small RNA
that mediates a protective response against inner membrane
damage (Chao and Vogel, 2016). Various types of cellular
stress trigger global shifts in translation. In yeast, formation of
the prion [PSI+] form of the termination factor protein Sup35p
promotes read-through of nonsense codons, providing a sur-
vival advantage under various conditions (True et al., 2004).
Analogously, the induction of 30 UTR translation under oxidative
stress in aging or neurodegenerative disease might contribute
in some way to the stress response.STAR+METHODS
Detailed methods are provided in the online version of this paper
and include the following:
d KEY RESOURCES TABLE
d CONTACT FOR REAGENT AND RESOURCE SHARING
d EXPERIMENTAL MODEL AND SUBJECT DETAILSB Mouse models
B Cell Lines
d METHOD DETAILS
B TRAP profiling
B Datasets
B RNaseq / TRAPseq read mapping and expression
quantification
B Calculation of Rtc
B Meta analyses
B Genes correlated with Rtc
B Sequence Logos
B Lipofuscin quantification
B Western Blotting
B Ribosome Profiling
B TEX Treatment
B Analysis of proteomic data
d QUANTIFICATION AND STATISTICAL ANALYSIS
d DATA AND SOFWARE AVAILABILITYSUPPLEMENTAL INFORMATION
Supplemental Information includes five figures and four tables and can be
found with this article online at https://doi.org/10.1016/j.celrep.2018.10.094.
ACKNOWLEDGMENTS
We thank Phil Sharp, Torben Jensen, Søren Lykke-Andersen, Wendy Gilbert,
Marcus E. Raichle, Manu S. Goyal, and members of the Burge lab for com-
ments on the manuscript and Hiten Madhani for helpful suggestions regarding
iron-sulfur cluster proteins. P.H.S. was supported by a Genentech Life Sci-
ences Research Fellowship, and H.L. was supported by a postdoctoral fellow-
ship from the JPB Foundation. This work was supported by an award from the
JPB Foundation (M.H.) and by NIH grant number HG002439 (to C.B.B.).
AUTHOR CONTRIBUTIONS
P.H.S. performed experiments, analyses, and drafted the manuscript. H.L.
performed mouse microscopy and additional experiments. D.D. contributed
to experiments. M.S.G. and M.E.R. contributed PET experimental data and
analysis of brain metabolism. M.H. supervised the project, performed TRAP
experiments, and revised the manuscript. C.B.B. supervised the project and
revised the manuscript. All authors reviewed and approved the final
manuscript.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: June 26, 2017
Revised: September 24, 2018
Accepted: October 25, 2018
Published: November 27, 2018
REFERENCES
Alhebshi, A., Sideri, T.C., Holland, S.L., and Avery, S.V. (2012). The essential
iron-sulfur protein Rli1 is an important target accounting for inhibition of cell
growth by reactive oxygen species. Mol. Biol. Cell 23, 3582–3590.
Arribere, J.A., and Fire, A.Z. (2018). Nonsense mRNA suppression via nonstop
decay. eLife 7, 1470.
Berkovits, B.D., and Mayr, C. (2015). Alternative 30 UTRs act as scaffolds to
regulate membrane protein localization. Nature 522, 363–367.
Brumaghim, J.L., Li, Y., Henle, E., and Linn, S. (2003). Effects of hydrogen
peroxide upon nicotinamide nucleotide metabolism in Escherichia coli:
changes in enzyme levels and nicotinamide nucleotide pools and studies of
the oxidation of NAD(P)H by Fe(III). J. Biol. Chem. 278, 42495–42504.
Carlyle, B.C., Kitchen, R.R., Kanyo, J.E., Voss, E.Z., Pletikos, M., Sousa,
A.M.M., Lam, T.T., Gerstein, M.B., S
estan, N., and Nairn, A.C. (2017). A multi-
regional proteomic survey of the postnatal human brain. Nat. Neurosci. 20,
1787–1795.
Carninci, P., Sandelin, A., Lenhard, B., Katayama, S., Shimokawa, K., Pon-
javic, J., Semple, C.A.M., Taylor, M.S., Engström, P.G., Frith, M.C., et al.
(2006). Genome-wide analysis of mammalian promoter architecture and evo-
lution. Nat. Genet. 38, 626–635.
Carpenter, A.E., Jones,T.R., Lamprecht,M.R.,Clarke,C., Kang, I.H., Friman,O.,
Guertin, D.A., Chang, J.H., Lindquist, R.A., Moffat, J., et al. (2006). CellProfiler:
image analysis software for identifying and quantifying cell phenotypes.
Genome Biol. 7, R100.
Chao, Y., and Vogel, J. (2016). A 30 UTR-derived small RNA provides the reg-
ulatory noncoding arm of the inner membrane stress response. Mol. Cell 61,
352–363.
Corral-Debrinski, M., Horton, T., Lott, M.T., Shoffner, J.M., Beal, M.F., and
Wallace, D.C. (1992). Mitochondrial DNA deletions in human brain: regional
variability and increase with advanced age. Nat. Genet. 2, 324–329.Cell Reports 25, 2447–2456, November 27, 2018 2455
Dabrowski, M., Bukowy-Bieryllo, Z., and Zietkiewicz, E. (2015). Translational
readthrough potential of natural termination codons in eucaryotes–The impact
of RNA sequence. RNA Biol. 12, 950–958.
Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut,
P., Chaisson,M., andGingeras, T.R. (2013). STAR: ultrafast universal RNA-seq
aligner. Bioinformatics 29, 15–21.
Doma, M.K., and Parker, R. (2006). Endonucleolytic cleavage of eukaryotic
mRNAs with stalls in translation elongation. Nature 440, 561–564.
Fejes-Toth, K., Sotirova, V., Sachidanandam, R., and Assaf, G. (2009). Post-
transcriptional processing generates a diversity of 50-modified long and short
RNAs. Nature 457, 1028–1032.
Garneau, N.L., Wilusz, J., andWilusz, C.J. (2007). The highways and byways of
mRNA decay. Nat. Rev. Mol. Cell Biol. 8, 113–126.
Gerashchenko, M.V., and Lobanov, A.V. (2012). Genome-wide ribosome
profiling reveals complex translational regulation in response to oxidative
stress. Proc. Natl. Acad. Sci. USA 109, 17394–17399.
Gokce, O., Stanley, G.M., Treutlein, B., Neff, N.F., Camp, J.G., Malenka, R.C.,
Rothwell, P.E., Fuccillo, M.V., Su€dhof, T.C., and Quake, S.R. (2016). cellular
taxonomy of the mouse striatum as revealed by single-cell RNA-seq. Cell
Rep. 16, 1126–1137.
Goodwin, M., Mohan, A., Batra, R., Lee, K.-Y., Charizanis, K., Fernández Gó-
mez, F.J., Eddarkaoui, S., Sergeant, N., Buée, L., Kimura, T., et al. (2015).
MBNL sequestration by toxic RNAs and RNA misprocessing in the myotonic
dystrophy brain. Cell Rep. 12, 1159–1168.
Goyal, M.S., Vlassenko, A.G., Blazey, T.M., Su, Y., Couture, L.E., Durbin, T.J.,
Bateman, R.J., Benzinger, T.L.-S., Morris, J.C., and Raichle, M.E. (2017). Loss
of brain aerobic glycolysis in normal human aging. Cell Metab. 26, 353–360.e3.
GTEx Consortium (2015). Human genomics. The Genotype-Tissue Expression
(GTEx) pilot analysis: multitissue gene regulation in humans. Science 348,
648–660.
Guydosh, N.R., and Green, R. (2014). Dom34 rescues ribosomes in 30 untrans-
lated regions. Cell 156, 950–962.
Hansen, T.B., Jensen, T.I., Clausen, B.H., Bramsen, J.B., Finsen, B., Dam-
gaard, C.K., and Kjems, J. (2013). Natural RNA circles function as efficient
microRNA sponges. Nature 495, 384–388.
Heiman, M., Schaefer, A., Gong, S., Peterson, J.D., Day, M., Ramsey, K.E.,
Suárez-Fariñas, M., Schwarz, C., Stephan, D.A., Surmeier, D.J., et al. (2008).
A translational profiling approach for the molecular characterization of CNS
cell types. Cell 135, 738–748.
Heiman, M., Kulicke, R., Fenster, R.J., Greengard, P., and Heintz, N. (2014).
Cell type-specific mRNA purification by translating ribosome affinity purifica-
tion (TRAP). Nat. Protoc. 9, 1282–1291.
Hornstein, N., Torres, D., Das Sharma, S., Tang, G., Canoll, P., and Sims, P.A.
(2016). Ligation-free ribosome profiling of cell type-specific translation in the
brain. Genome Biol. 17, 149.
Ibrahim, F., Maragkakis, M., Alexiou, P., and Mourelatos, Z. (2018). Ribothryp-
sis, a novel process of canonical mRNA decay, mediates ribosome-phased
mRNA endonucleolysis. Nat. Struct. Mol. Biol. 25, 302–310.
Karginov, F.V., Cheloufi, S., Chong, M.M.W., Stark, A., Smith, A.D., and Han-
non, G.J. (2010). Diverse endonucleolytic cleavage sites in the mammalian
transcriptome depend upon microRNAs, Drosha, and additional nucleases.
Mol. Cell 38, 781–788.
Kocabas, A., Duarte, T., Kumar, S., and Hynes, M.A. (2015). Widespread differ-
ential expression of coding region and 30 UTR sequences in neurons and other
tissues. Neuron 88, 1149–1156.
Kuersten, S., and Goodwin, E.B. (2003). The power of the 30 UTR: translational
control and development. Nat. Rev. Genet. 4, 626–637.
Leng, N., Dawson, J.A., Thomson, J.A., Ruotti, V., Rissman, A.I., Smits,
B.M.G., Haag, J.D., Gould, M.N., Stewart, R.M., and Kendziorski, C. (2013).
EBSeq: an empirical Bayes hierarchical model for inference in RNA-seq exper-
iments. Bioinformatics 29, 1035–1043.2456 Cell Reports 25, 2447–2456, November 27, 2018Li, B., and Dewey, C.N. (2011). RSEM: accurate transcript quantification
from RNA-seq data with or without a reference genome. BMC Bioinformatics
12, 323.
López-Otı́n, C., Blasco, M.A., Partridge, L., Serrano, M., and Kroemer, G.
(2013). The hallmarks of aging. Cell 153, 1194–1217.
Malka, Y., Steiman-Shimony, A., Rosenthal, E., Argaman, L., Cohen-Daniel, L.,
Arbib, E., Margalit, H., Kaplan, T., and Berger, M. (2017). Post-transcriptional
30-UTR cleavage of mRNA transcripts generates thousands of stable uncap-
ped autonomous RNA fragments. Nat. Commun. 8, 2029.
Martin, M. (2011). Cutadapt removes adapter sequences from high-throughput
sequencing reads. EMBnet.Journal 17. https://doi.org/10.14806/ej.17.1.200.
Mattson, M.P., and Magnus, T. (2006). Ageing and neuronal vulnerability. Nat.
Rev. Neurosci. 7, 278–294.
McIlwain, S., Tamura, K., Kertesz-Farkas, A., Grant, C.E., Diament, B., Frewen,
B., Howbert, J.J., Hoopmann, M.R., Käll, L., Eng, J.K., et al. (2014). Crux: rapid
open source protein tandem mass spectrometry analysis. J. Proteome Res.
13, 4488–4491.
Mercer, T.R., Wilhelm, D., Dinger, M.E., Soldà, G., Korbie, D.J., Glazov, E.A.,
Truong, V., Schwenke, M., Simons, C., Matthaei, K.I., et al. (2011). Expression
of distinct RNAs from 30 untranslated regions. Nucleic Acids Res. 39, 2393–
2403.
Mills, E.W., Wangen, J., Green, R., and Ingolia, N.T. (2016). Dynamic regula-
tion of a ribosome rescue pathway in erythroid cells and platelets. Cell Rep.
17, 1–10.
Nandy, K. (1971). Properties of neuronal lipofuscin pigment in mice. Acta Neu-
ropathol. 19, 25–32.
Puig, S., Askeland, E., and Thiele, D.J. (2005). Coordinated remodeling of
cellular metabolism during iron deficiency through targeted mRNA degrada-
tion. Cell 120, 99–110.
Starck, S.R., Tsai, J.C., Chen, K., Shodiya, M., Wang, L., Yahiro, K., Martins-
Green, M., Shastri, N., and Walter, P. (2016). Translation from the 50 untrans-
lated region shapes the integrated stress response. Science 351, aad3867.
True, H.L., Berlin, I., and Lindquist, S.L. (2004). Epigenetic regulation of trans-
lation reveals hidden genetic variation to produce complex traits. Nature 431,
184–187.
Tsuboi, T., Kuroha, K., Kudo, K., Makino, S., Inoue, E., Kashima, I., and Inada,
T. (2012). Dom34:Hbs1 plays a general role in quality-control systems by
dissociation of a stalled ribosome at the 30 end of aberrant mRNA. Mol. Cell
46, 518–529.
Veatch, J.R., McMurray, M.A., Nelson, Z.W., and Gottschling, D.E. (2009).
Mitochondrial dysfunction leads to nuclear genome instability via an iron-sulfur
cluster defect. Cell 137, 1247–1258.
Vlassenko, A.G., Vaishnavi, S.N., Couture, L., Sacco, D., Shannon, B.J., Mach,
R.H., Morris, J.C., Raichle, M.E., and Mintun, M.A. (2010). Spatial correlation
between brain aerobic glycolysis and amyloid-b (Ab ) deposition. Proc. Natl.
Acad. Sci. USA 107, 17763–17767.
Wu, X., and Bartel, D.P. (2017). kpLogo: positional k-mer analysis reveals hid-
den specificity in biological sequences. Nucleic Acids Res. 45 (W1), W534–
W538.
Young, D.J., Guydosh, N.R., Zhang, F., Hinnebusch, A.G., and Green, R.
(2015). Rli1/ABCE1 recycles terminating ribosomes and controls translation
reinitiation in 3’UTRs in vivo. Cell 162, 872–884.
Zhang, X.O., Wang, H.B., Zhang, Y., Lu, X., Chen, L.L., and Yang, L. (2014).
Complementary sequence-mediated exon circularization. Cell 159, 134–147.
Zhang, Y., Sloan, S.A., Clarke, L.E., Caneda, C., Plaza, C.A., Blumenthal,
P.D., Vogel, H., Steinberg, G.K., Edwards, M.S.B., Li, G., et al. (2016). Puri-
fication and characterization of progenitor and mature human astrocytes re-
veals transcriptional and functional differences with mouse. Neuron 89,
37–53.
Zou, X., Ratti, B.A., O’Brien, J.G., Lautenschlager, S.O., Gius, D.R., Bonini,
M.G., and Zhu, Y. (2017). Manganese superoxide dismutase (SOD2): is there
a center in the universe of mitochondrial redox signaling? J. Bioenerg. Bio-
membr. 49, 325–333.
STAR+METHODSKEY RESOURCES TABLEREAGENT or RESOURCE SOURCE IDENTIFIER
Antibodies
Rabbit monoclonal anti-ABCE1 Abcam ab185548
Rabbit polyclonal anti-Histone H3 Abcam ab1791; RRID: AB_302613
Rabbit polyclonal anti-GFP Abcam ab6556; RRID: AB_305564
Rabbit monoclonal anti-EXOSC3 Abcam ab190689
Mouse monoclonal anti-b-actin Sigma A5441; RRID: AB_476744
Biological Samples
DMEM ThermoFisher 11965-092
Opti-MEM ThermoFisher 31985-092
FBS Hyclone SH30071.03
Human Origene TissueScan cDNA Origene Hbrt101
Mouse Origene TissueScan cDNA Origene Mdrt101
Chemicals, Peptides, and Recombinant Proteins
Lipofectamine RNAiMAX ThermoFisher 13778150
phosphatase inhibitor cocktail ThermoFisher 78441
BCA ThermoFisher 23227
ECL Plus ThermoFisher 32132
Critical Commercial Assays
QIAGEN RNeasy spin columns QIAGEN 74106
On-column DNase Digestion QIAGEN 79254
Hi Capacity cDNA kit ThermoFisher 4368813
KAPA Sybr Master Mix Kapa Biosystems KK4618
TaqMan Gene Expression Master Mix ThermoFisher 4369514
Deposited Data
TRAP sequencing of young and aged D1 and D2 SPNs GEO GSE97461
Ribosome Profiling under different conditions GEO GSE97461
Experimental Models: Cell Lines
NIH 3T3 cell line ATCC Cat# CRL-6442
Experimental Models: Organisms/Strains
Drd1::EGFP-L10a (C56BL/6J background) Laboratory of M. Heiman Drd1::EGFP-L10a
Drd2::EGFP-L10a (C56BL/6J background) Laboratory of M. Heiman Drd2::EGFP-L10a
Oligonucleotides
See Table S2 This Study N/A
Sequence-Based Reagents
Silencer Select pre-designed siRNA Abce1 siRNA ThermoFisher s76800
Silencer Select pre-designed siRNA negative control siRNA ThermoFisher 4390844
Silencer Select pre-designed siRNA Exosc3 siRNA ThermoFisher s83101
TaqMan Human ABCE1 primer ThermoFisher Hs01003006_g1 FAM-MGB
TaqMan Human ACTB primer ThermoFisher Hs01060655_g1 VIC-MGB-PL
TaqMan Mouse Abce1 primer ThermoFisher Mm00649858_m1 FAM-MGB
TaqMan Mouse Actb primer ThermoFisher Mm00607939_s1 VIC-MGB-PL
(Continued on next page)
Cell Reports 25, 2447–2456.e1–e4, November 27, 2018 e1
Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
Software and Algorithms
STAR version 2.5.1b (Dobin et al., 2013) https://github.com/alexdobin/STAR
EBSeq version 1.1.5 (Leng et al., 2013) https://bioconductor.org/packages/release/bioc/
html/EBSeq.html
RSEM 1.2.20 (Li and Dewey, 2011) https://deweylab.github.io/RSEM/
CIRCexplorer 1.1.10 (Zhang et al., 2014) https://github.com/YangLab/CIRCexplorer
CellProfiler (Carpenter et al., 2006) http://cellprofiler.org
FASTX toolkit 0.0.14 http://hannonlab.cshl.edu/fastx_toolkit/index.html
Cutadapt version 1.14 (Martin, 2011) http://cutadapt.readthedocs.io/en/stable/
installation.html
Other
RNA Sequencing Datasets Assessed – See Table S1 N/A N/ACONTACT FOR REAGENT AND RESOURCE SHARING
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact,
Christopher Burge (cburge@mit.edu).
EXPERIMENTAL MODEL AND SUBJECT DETAILS
All mouse experiments were conducted with the approval and oversight of the MIT Animal Care and Use Committee.
Mouse models
Female BAC transgenic mouse lines Drd1::EGFP-L10a and Drd2::EGFP-L10a on a C56BL/6J background at 6 weeks of age (PN42)
and 2-years of age were used for experiments. Mice were decapitated and brain tissue was immediately dissected and used for
TRAP RNA purification as described in Heiman et al., 2008 (Heiman et al., 2008).
Cell Lines
NIH 3T3 cells were grown at 37C at 5% CO2 in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin. Knockdowns
were performed with Silencer Select pre-designed siRNAs. Transfection was performed with Lipofectamine RNAiMAX Reagent in
Opti-MEM as per manufactures instructions with 75pmol of siRNA per well of a 6-well cell-culture dish for ABCE1 knockdown
and 400pmol for EXOSC3 knockdown.
METHOD DETAILS
TRAP profiling
Immediately following decapitation, TRAP profiling was performed as detailed in Heiman et al., 2014 (Heiman et al., 2014). RNA-seq
libraries were prepped using NuGEN Ovation V2.0 system.
Datasets
Datasets assessed in this study and associated accession numbers and PubMed IDs if available are listed in Table S1.
RNaseq / TRAPseq read mapping and expression quantification
Reads were mapped to the mouse or human genomes version GRCm38/mm10 or GRCh37/hg19 respectively using the STAR
read aligner version 2.5.1b (Dobin et al., 2013) with ENSEMBL Version 75 gene annotations to guide exon-exon junction align-
ment. Gene expression values were quantified using RSEM version 1.2.20 (Li and Dewey, 2011) and differences in gene expres-
sion between experimental conditions were estimated using EBSeq version 1.1.5 (Leng et al., 2013) again using ENSEMBL
Version 75 annotations. To identify junction reads, which might indicate the presence of circular RNAs, we used the chimeric/
fusion alignment functionality of STAR. Reads were mapped using standard parameters in addition to: ‘‘–chimSegmentMin
15–chimJunctionOverhangMin 15.’’ The CIRCexplorer tool was then used to quantify circular RNAs from these alignments
(Zhang et al., 2014).e2 Cell Reports 25, 2447–2456.e1–e4, November 27, 2018
Calculation of Rtc
Rtc, or the ‘‘termination codon ratio,’’ was defined as the log base 2 of the ratio of the mean read coverage after the annotated stop
codon to the mean read coverage before the annotated stop codon: !
1 Pi = l
 i = tccvgðiÞRtc = log l tc2 1 Pk = tc 1
 k = 0 cvgðkÞtc 1
assuming a gene of length l with an annotated stop codon at tc. A minimum value of 1 was used for numerator and denominator to
avoid division by zero. Only constitutive exons were included in the calculation and overlapping genes and genes with multiple
annotated protein coding stop codons were excluded from analyses. Negative Rtc values occur when the mean 3
0 UTR coverage
is less than the mean gene body coverage, since log(x) < 0 when x < 1. We have found that for some library preparations sequence
coverage over the 30 UTR is, on average, lower than that over the gene body across most genes, leading to negative Rtc values. In
general, because of the potential for such biases, Rtc can be compared only between different experimental conditions or treatments
for which library preparation and sequencing did not vary. When assayed by qRT-PCR, genes with high Rtc values generally had
30 UTR expression that was severalfold higher than the associated CDS, indicating that CDS-containing transcripts from these
loci are typically reduced but not absent.
Meta analyses
Meta-analysis plots of gene RNA-seq or ribosome profiling coverage were performed using constitutive exons of non-overlapping
genes and excluded genes with multiple annotated protein coding stop codons. To plot the heatmap, coverage upstream and
downstream of the stop codon (the 50 UTR + ORF, and the 30 UTR respectively), were normalized to windows of 1000 bins each.
The minimum window coverage per transcript was subtracted from all windows and the result was normalized to sum to 1, log
transformed, and smoothed using a Gaussian kernel 100 windows wide using the smth function of the R smoother package.
Genes correlated with Rtc
To identify genes with expression patterns correlated with Rtc, we first quantified the mean per-sample Rtc. We then computed the
Pearson correlation of this vector with the vector of gene expression values for each of 18745 nuclear protein-coding genes supple-
mented with all genes encoded on the mitochondrial chromosome. The distribution of P values is plotted in Figure S4E. We then
ranked the 3938 positively correlated genes by their Pearson correlation coefficients to identify genes expression patterns that
accompanied increased Rtc.
Sequence Logos
Sequence logos in Figure S5A were generated with kpLogo (Wu and Bartel, 2017) from stop codon centered kmers ranked by Rtc.
Lipofuscin quantification
Male mice aged 9 weeks or 19 months from the Drd1::EGFP-L10a orDrd2::EGFP-L10a Bacterial Artificial Chromosome (BAC) trans-
genic lines (n = 3 each group) were used for lipofuscin measurements. Mouse brain tissue was perfused, embedded, fixed with 4%
paraformaldehyde and cryosectioned at 20 mm thickness. Drd1 or Drd2 neurons expressing EGFP-L10a were stained with anti-GFP
antibody (Abcam ab6556 1:5,000 dilution). Autofluorescent lipofuscin was visualized in both red and far-red channel. Zeiss LSM700
Confocal Microscope with a 40X objective lens was used. The percentage of Drd1 or Drd2 neurons containing lipofuscin was
obtained by an examination of three tissue sections per mouse where two striatal images were taken in each tissue section. The
proportion of lipofuscin per cell was quantified using CellProfiler (Carpenter et al., 2006).
Western Blotting
Cells were lysed with 1%SDS in water containing protease and phosphatase inhibitor cocktail (ThermoFisher). Protein concentration
wasmeasuredwith BCA (ThermoFisher) and loaded at 20 mg per lane. After electrophoresis, the proteinswere transferred semi-dry to
PVDFmembrane. Themembranewas incubated with 5%milk in PBS-T (PBSwith 0.05%Tween20) for one hour at room temperature
then incubated with following antibodies for over-night at 4C: Abce1 (Abcam ab185548, 1:1000 dilution), Exosc3 (Abcam ab190689,
1:1000 dilution), Histone H3 (Abcam ab1791, 1:20,000 dilution). The following day, the membrane was washed three-times with
PBS-T for 10 minutes each, incubated with a HRP-conjugated anti-rabbit secondary antibody for one hour at room temperature
(1:10,000 dilution in PBS-T), washed three times, and developed with ECL Plus (ThermoFisher).
Ribosome Profiling
Ribosome profiling was performed using the Illumina TruSeq Ribo Profile kit (RPHMR12126) according to manufacturer’s specifica-
tions with the following exceptions. Footprints spanning 15-32nt were excised from the RNA gel to capture both short and long
ribosome protected fragments. Following end repair of size selected RNA fragments, sequencing libraries were prepped using
the Clontech SMARTer smRNA-Seq kit (635030) as described in Hornstein et al. (Hornstein et al., 2016). Libraries were multiplexCell Reports 25, 2447–2456.e1–e4, November 27, 2018 e3
sequenced on an Illumina NextSeq and the resulting demultiplexed sequences were collapsed to unique reads using fastx_collapser
and trimmed of the polyA tails, template switching nucleotides, and adaptor sequences added by the SMARTer smRNA-Seq
protocol using cutadapt.
TEX Treatment
TEX treatments were performed as described in Malka et al. (Malka et al., 2017).
Analysis of proteomic data
Mass spectrometry data for PXD001250 were downloaded from the PRIDE archive and RAW files were converted to mzML using
MSconvert from the proteowizard package (http://proteowizard.sourceforge.net/). A custom peptide database was constructed
by translating all 30 UTR open reading frames 18 amino acids or longer. This database was combined with the UniProt SwissProt
and trEMBL databases. The CRUX (McIlwain et al., 2014) mass spectrometry analysis toolkit was used to perform all subsequent
searching, ranking and quantification of proteins and peptides and to construct a decoy database of peptides to empirically calculate
a false discovery rate. The decoy database was constructed using the CRUX amino acid permutation method. Peptides were
searched against our custom database and the decoy database using comet allowing for fixed carbamidomethylation modifications
and variable N-acetylation and methionine oxidation. Percolator was then used to rank and score proteins using default options in
addition to ‘‘–protein T’’ to perform fido protein scoring and ranking. A q-value threshold of 0.01 was used to filter the resulting
proteins/peptides. Only 30 UTR peptides that could be unambiguously assigned to having originated from the 30 UTR, and not canon-
ical ORFs, were considered in downstream analysis.
QUANTIFICATION AND STATISTICAL ANALYSIS
Statistical parameters and specific tests are reported in the text and figures including the statistical significance. Significance was
defined as p < 0.01. All statistical test were performed using the R programming language. No statistical methods were employed
a-priori for randomization, stratification, sample size estimation, or exclusion of any samples.
DATA AND SOFWARE AVAILABILITY
Sequencing datasets generated in this study have been deposited under accession number GEO: GSE97461. All software used in
this manuscript is referenced in the methods. All custom code and pipelines are available upon request.e4 Cell Reports 25, 2447–2456.e1–e4, November 27, 2018