Identification of residues within the L2 region of rhesus TRIM5alpha that are required for retroviral restriction and cytoplasmic body localization.

Abstract

The intracellular restriction factor TRIM5alpha, inhibits infection by numerous retroviruses in a species-specific manner. The best characterized example of this restriction is the TRIM5alpha protein from rhesus macaques (rhTRIM5alpha), which potently inhibits HIV-1 infection. TRIM5alpha localizes to cytoplasmic assemblies of protein referred to as cytoplasmic bodies, though the role that these bodies play in retroviral restriction is unclear. We employed a series of truncation mutants to identify a discrete region, located within the Linker2 region connecting the coiled-coil and B30.2/PRYSPRY domains of TRIM5alpha, which is required for cytoplasmic body localization. Deletion of this region in the context of full-length rhTRIM5alpha abrogates cytoplasmic body localization. Alanine mutagenesis of the residues in this region identifies two stretches of amino acids that are required for both cytoplasmic body localization and retroviral restriction. This work suggests that the determinants that mediate TRIM5alpha localization to cytoplasmic bodies play a requisite role in retroviral restriction.

Full text

Page 1

Identification of residues within the L2 region of rhesus TRIM5α that are required for
retroviral restriction and cytoplasmic body localization
Jaya Sastric, Christopher O'Connorc, Cindy M. Danielsona, Michael McRavena, Patricio Perezb,
Felipe Diaz-Grifferob, Edward M. Campbellc,⁎
aDepartment of Cell and Molecular Biology, Feinberg School of Medicine, Northwestern University, Chicago IL, USA
bDepartment of Microbiology and Immunology, Albert Einstein College of Medicine, Bronx New York, USA
cDepartment of Microbiology and Immunology, Stritch School of Medicine, Loyola University Chicago, Maywood IL, USA
a b s t r a c ta r t i c l ei n f o
Article history:
Received 22 April 2010
Returned to author for revision 15 May 2010
Accepted 7 June 2010
Available online 14 July 2010
Keywords:
TRIM5α
Multimerization
TRIM
HIV-1
Restriction factor
The intracellular restriction factor TRIM5α, inhibits infection by numerous retroviruses in a species-specific
manner. The best characterized example of this restriction is the TRIM5α protein from rhesus macaques
(rhTRIM5α), which potently inhibits HIV-1 infection. TRIM5α localizes to cytoplasmic assemblies of protein
referred to as cytoplasmic bodies, though the role that these bodies play in retroviral restriction is unclear.
We employed a series of truncation mutants to identify a discrete region, located within the Linker2 region
connecting the coiled-coil and B30.2/PRYSPRY domains of TRIM5α, which is required for cytoplasmic body
localization. Deletion of this region in the context of full-length rhTRIM5α abrogates cytoplasmic body
localization. Alanine mutagenesis of the residues in this region identifies two stretches of amino acids that
are required for both cytoplasmic body localization and retroviral restriction. This work suggests that the
determinants that mediate TRIM5α localization to cytoplasmic bodies play a requisite role in retroviral
restriction.
© 2010 Elsevier Inc. All rights reserved.
Introduction
The species-specific tropism of numerous retroviruses is deter-
mined by host cell proteins, termed restriction factors, which inhibit
viral replication at numerous stages of the viral life cycle. The TRIM
family of proteins represents one such class of restriction factors, of
which the best characterized example is the ability of TRIM5α from
rhesus macaques (rhTRIM5α) to inhibit human immunodeficiency
virus type-1 (HIV-1) (Stremlau et al., 2004).
TRIM proteins, including TRIM5α, are defined by a TRIpartite
Motif, which includes an N-terminal RING domain, B-Box/B-Box2
domain and coiled-coil domains (Reymond et al., 2001). The RING
domain contains a putative E3 ubiquitin ligase activity, and the
established biological activity of many TRIM proteins has been shown
to require the E3 ligase activity mediated by this domain (Balastik
et al., 2008; Barr et al., 2008; Gack et al., 2007; Meroni and Diez-Roux,
2005; Ozato et al., 2008). The B-Box 2 domain has been shown to be
important in mediating the cooperative, higher-order multimeriza-
tion of TRIM5α during restriction (Diaz-Griffero et al., 2007, 2009,
2006). The coiled-coil domain has been shown to mediate the low-
order multimerization of many proteins, including TRIM family
proteins. In the case of TRIM5α, this is currently thought to facilitate
protein dimerization (Kar et al., 2008; Langelier et al., 2008). In
addition to these domains, TRIM5α also contains a C-terminal B30.2/
PRYSPRY (hereafter SPRY) domain, which has been shown to dictate
the species-specific antiviral activity of TRIM5α proteins (Sawyer
et al., 2005; Song et al., 2005b; Yap et al., 2005). In some cases, most
notably that of owl monkeys, the SPRY domain has been functionally
replaced by the retrotransposition of cyclophilin A into the TRIM5
locus (Brennan et al., 2008; Newman et al., 2008; Sayah et al., 2004;
Wilson et al., 2008). The resulting TRIM-Cyp protein also exhibits
antiviral activity against various retroviruses.
Like most other TRIM family proteins (Reymond et al., 2001),
TRIM5α is observed to localize to cytoplasmic accumulations of
protein termed cytoplasmic bodies (Campbell et al., 2007; Stremlau
et al., 2004). However, the role that these cytoplasmic bodies play in
the process of restriction remains unclear. Two studies have found
that pre-existing cytoplasmic bodies are not required for restriction
(Perez-Caballero et al., 2005; Song et al., 2005a). In one of these
studies, the authors observe that a cell line stably expressing TRIM-
Cyp can potently restrict HIV-1 infection in the absence of notable
cytoplasmic bodies in these cells (Perez-Caballero et al., 2005). In the
second study, Song and colleagues observed that treatment of cells
stably expressing rhTRIM5α with the heat shock protein inhibitor
Geldanamycin, reduced or eliminated the localization of TRIM5α to
cytoplasmic bodies without affecting the ability of rhTRIM5α to
restrict HIV-1 infection (Song et al., 2005a). These studies collectively
Virology 405 (2010) 259–266
⁎ Corresponding author. 2160 S. First Ave, Maywood, IL 60153, USA.
E-mail address: ecampbell@lumc.eduU (E.M. Campbell).
0042-6822/$ – see front matter © 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.virol.2010.06.015
Contents lists available at ScienceDirect
Virology
journal homepage: www.elsevier.com/locate/yviro

Page 2

suggest that cytoplasmic body localization is not required for retro-
viral restriction.
Alternatively, our previous studies have observed that cytoplasmic
bodies are dynamic structures which turnover rapidly and can traffic
through the cell utilizing the microtubule network (Campbell et al.,
2007). This study also demonstrated that TRIM5α exists in two
populations in cells, including a population localized to cytoplasmic
bodies and a more diffuse cytoplasmic population (Campbell et al.,
2007). We have also utilized fluorescently labeled HIV-1 virions to
observe virions associating with rhTRIM5α cytoplasmic bodies in
restricted cells. In this study, live cell imaging was able to visualize the
de novo formation of cytoplasmic bodies around individual, fluores-
cently labeled HIV-1 virions (Campbell et al., 2008). These studies
suggest that cytoplasmic bodies might play an important role in the
interactions occurring between TRIM5α and incoming retroviral
capsids during restriction.
We reasoned that if we could identify TRIM5α mutants that fail to
localize to cytoplasmic bodies, we could determine if the ability to
restrict retroviral infection is genetically separable from the ability to
localize to cytoplasmic bodies. In this study, we utilized a series of
truncationmutants to identify regions of TRIM5α that are required for
cytoplasmic body localization of TRIM5α. These truncation mutants
identified the region between the coiled-coil and the SPRY domains as
having determinants that mediate the accumulation of TRIM5α in
cytoplasmic bodies, previously described as the linker 2 region
(Javanbakht et al., 2006). Alanine scanning mutagenesis of this region
identified two distinct regions of the L2 domain that are required for
cytoplasmic body localization. Disruption of either of these regions
also generated TRIM5α proteins that lost the ability to restrict
infection. Alternatively, similar mutations within this region that did
not disrupt rhTRIM5α localization to cytoplasmic bodies did not
impair the ability of rhTRIM5α to restrict retroviral infection. This
suggests that the L2 region of TRIM5α is relevant to the ability of
TRIM5α to self-associate, allowing TRIM5α proteins to form multi-
meric assemblies around a virion, which is likely to be critical during
the restriction process.
Results
Identification of a region of the Linker2 region of rhTRIM5α required for
cytoplasmic body localization
To understand the determinants that influence the cellular
localization of rhTRIM5α, we generated a series of GFP fusion proteins
to regions of rhTRIM5α that have been implicated in facilitating the
multimerization or high-order multimerization of rhTRIM5α
(Fig. 1A). GFP fusions to the coiled-coil domain alone (data not
shown) and to the B-Box2 and coiled-coil domains (amino acids 84–
236) exhibited a diffuse localization in both the cytoplasm and the
nucleus. In contrast, GFP fusions including the B-Box2, coiled-coil
domain and L2 region (84–300) formed discrete accumulations in the
cytoplasm and the nucleus (Supplemental Figure 1), as did a similar
construct including the SPRY domain (84-end). This suggests that the
L2 region of rhTRIM5α possesses determinants required for cytoplas-
mic body formation.
To further our understanding of which residues in this region are
requiredfor cytoplasmic body localization ofrhTRIM5α, wegenerated
a series of C-terminal truncations within the L2 region (Fig. 1B). All
C-terminal truncations that included amino acids 263–278 promi-
nently localized to discrete, cytoplasmic accumulations that resem-
bled the cytoplasmic bodies formedby full-length rhTRIM5α (Fig. 1B).
In contrast, truncations including amino acids 1–233 or 1–248, and 1–
263 exhibited a diffuse localization in transfected cells. This result
suggests that the amino acids 263–278 include determinants required
for rhTRIM5α cytoplasmic body formation.
Two discrete regions of L2 are required for rhTRIM5α cytoplasmic body
localization
We next sought to identify which specific residues within amino
acids 263–278 are required for cytoplasmic body localization. We
reasoned that charged residues were more likely to contribute to or
facilitate protein–protein interactions that may be required for
cytoplasmic body formation. We therefore performed charge-clus-
ter-to-alanine scanning mutagenesis of the region (Bass et al., 1991;
Cunningham and Wells, 1989; Sebastian et al., 2009). Clusters of 2
charged amino acids within a 3 residue stretch were converted to
alanine, as shown in Fig. 2. Cell lines stably expressing YFP-tagged
forms of these proteins were generated and cell lines that expressed
comparable amounts of protein were chosen for subsequent analysis
(Fig. 2B). The localization of these YFP-TRIM5α fusions was examined
by fluorescence microscopy. As previously described, cytoplasmic
bodies were observable in HeLa cells expressing wild type YFP-
rhTRIM5α (Campbell et al., 2007) (Fig. 2C). Alanine mutagenesis of
amino acids 263–278 revealed that there are two discrete stretches of
the L2 region that are required for cytoplasmic body localization. The
first includes the residues 266–268, as rhTRIM5α variant in which
these residues were convertedto alanine(KPK266–268AAA) assumed
a diffuse, cytoplasmic localization (Fig. 2C). The second includes
residues 275–277, as the RRV275–277AAA mutant exhibited a diffuse
localization (Fig. 2C). In contrast, alanine mutagenesis of the amino
acids between these two patches, specifically TFH269–271AAA and
HKN271–273AAA, result in cytoplasmic body localization of these
rhTRIM5α mutants (Fig. 2C). Similar results were obtained with
TRIM5α mutants possessing an HA-epitope tag (data not shown).
In our analysis, we observed that TRIM5α localization was not
completely homogeneous in all cells. To quantitatively compare the
localizations exhibited by these various mutants, we collected 20 or
moreZ-stack imagesof each cell line using deconvolution microscopy.
We then performed quantitative image analysis on our images to
identify cytoplasmic bodies in cells using fixed fluorescence and size
definitions (Supplemental Movie 1; Fig. 3A). This allowed for the
unbiased, automated quantification of the number of cytoplasmic
bodies per cell for each cell line. As shown in Fig. 3, automated image
analysis of these cell lines recapitulated the phenotypes identified by
visual inspection. Cytoplasmic bodies were readily detected in HeLa
cells expressing wild type rhTRIM5α, as well as the cell lines ex-
pressing TRIM proteins harboring the TFH269–271AAA and HKN271–
273AAA mutations (Fig. 3B). Conversely, virtually no cytoplasmic
bodies were identified in the RRV275–277AAA variant, and very few
cytoplasmic bodies were identified in the KPK266–268AAA mutant,
though this analysis observed a statistically significant increase rela-
tive to the RRV275–277AAA mutant (pb0.0001). This indicates that,
while the KPK266–268AAA mutant exhibits a reduced ability to
localize to cytoplasmic bodies, it appears to be slightly less diminished
in this regard than the RRV275–277AAA mutant.
Cytoplasmic body localization does not correlate to TRIM5α protein
turnover
To assess if the localization of our individual TRIM5α variants to
cytoplasmic bodies correlated to their relative stability in cells, we
analyzed protein turnover in cells following cyclohexamide treat-
ment. As shown in Fig. 3B, we observed that the rate of turnover of
two of our mutants, RRV275–277AAA and TFH269-271AAA, was
decreased relative to the turnover of wild type rhTRIM5α and the
other mutants in our study. If cytoplasmic body localization were
affecting the turnover of TRIM5α, this would predict that mutants
that do not localize to cytoplasmic bodies exhibit aberrant turnover
compared to the wild type protein. However, while one of our diffuse
mutants exhibits a decreased rate of turnover (RRV275–277AAA), the
other diffuse mutant (KPK266–268) appears to be degraded with wild
260
J. Sastri et al. / Virology 405 (2010) 259–266

Page 3

type kinetics. Conversely, one mutant that localized normally to
cytoplasmic body localization (TFH269–271AAA) exhibited increased
stability relative to wild type, while another mutant localizing to
cytoplasmic bodies (HKN271–273AAA) was degraded with kinetics
similar to the wild type protein. Thus, while two mutants in our panel
exhibit a longer half-life, this did not correlate to cytoplasmic body
localization.
rhTRIM5α variants that do not localize to cytoplasmic bodies retain the
ability to multimerize
Wenextsoughttoassessifthemutationswehadintroducedintothe
L2regionwere grossly affectingtheconformationorfoldingofTRIM5α.
Previous biochemical characterization of TRIM5α, and a related variant
possessing the RING domain of TRIM21 have identified low-order
multimers of TRIM5α following biochemical crosslinking with EGS or
glutaraldehyde(Javanbakhtet al., 2005; Kar et al.,2008; Langelier et al.,
2008; Mische et al., 2005). To assess if our mutations were affecting the
ability of the protein to multimerize, we performed biochemical
crosslinking of our panel of mutant TRIM5α proteins. Glutaraldehyde
crosslinking of YFP-rhTRIM5α with increasing concentrations of
glutaraldehyde revealed multimeric forms of the protein at higher
glutaraldehyde concentrations (Fig. 4). The lowest molecular weight at
which crosslinked protein could be observed was approximately
150 kDa, which likely represents dimeric forms of TRIM5α. However,
the predominant low-order multimer detected following crosslinking
was approximately 250 kDa. While some previous studies originally
identified this band as a trimer (Javanbakht et al., 2006; Mische et al.,
2005),subsequentstudieshavedeterminedthatthiscomplexislikelyto
beadimerexhibitinganomalouslyslowelectrophoreticmobility(Karet
Fig. 1. Cytoplasmic body determinants are present in the L2 region of TRIM5α. A. Domain structure of rhesus TRIM5α, including the residues demarcating the L2 region of
the protein. B. HeLa cells were transiently transfected with the indicated GFP-rhTRIM5α truncation mutant. Images provided show an individual focal plane of collected,
deconvolved Z-stack images.
261
J. Sastri et al. / Virology 405 (2010) 259–266

Page 4

al., 2008; Langelier et al., 2008). These results are consistent with those
studies. Similar results were observed in the cases of the TFH269–
271AAAand HKN271–273AAAmutants, whichlocalized tocytoplasmic
bodies, as well as for the RRV275–277AAA and KPK266–268AAA
mutants which do not localize to cytoplasmic bodies (Fig. 4). All of the
proteins in this study were also able to form higher-order multimers
withincreasingconcentrationsofglutaraldehyde(Fig.4).Similarresults
were observed with HA-tagged forms of these mutants (data not
shown). These data suggest that the reduced cytoplasmic body
localization observed for the RRV275–277AAA and KPK266–268AAA
mutantsisnotduetothesemutantsbeinggrosslymisfoldedorunableto
form low-order or higher-order multimers.
rhTRIM5α variants that do not localize to cytoplasmic bodies do not
restrict HIV-1 infection
To examine the correlation between cellular localization and
restriction activity, we assessed the ability of our rhTRIM5α variants
to restrict HIV-1 infection. Cell lines expressing the indicated HA- or
YFP-tagged rhTRIM5α variants were infected with serial dilutions of
HIV-1 virus expressing GFP following infection (Campbell et al., 2007;
Stremlau et al., 2004). In our cell lines expressingYFP-rhTRIM5α, both
diffuse mutants KPK266–268AAA and RRV275–277AAA were infec-
ted as readily as control HeLa cells (Fig. 5). In contrast, cell lines
expressingthe TFH269–271AAAor HKN271–273AAA variants,both of
Fig. 2. Subcellular localization of YFP-tagged WT and mutant rhTRIM5 in HeLa cells. (A) Residues 263–278 within the L2 region of rhTRIM5α are shown. Charged residues within this
stretch that wereconverted to alanine are indicated by light ordark boxes. (B) TRIM5α protein levels were analyzed by Western Blotanalysis. HeLa cells stably expressing YFP-WT or
YFP-rhTRIM5α variants were lysed and the cell lysates were separated by SDS-PAGE. WT and mutant TRIM5α proteins were detected by immunoblotting withα-GFP antibody. Actin
was used as a loading control. (C) These stable cells were also fixed and examined for the subcellular localization of the YFP-WT and mutant TRIM5α proteins. Results are
representative of 3 independent experiments.
262
J. Sastri et al. / Virology 405 (2010) 259–266

Page 5

which localize to cytoplasmic bodies, restricted HIV-1 infection to a
similar degree as cells expressing wild type rhTRIM5α (Fig. 5). Thus,
the ability to restrict infection correlates with the ability of the protein
to localize into cytoplasmic bodies in these cell lines.
A similar pattern of restriction was observed when we infected our
cell lines expressing HA-tagged rhTRIM5α variants (data not shown).
Collectively, these data suggest that mutations that disrupt the ability
of rhTRIM5α to localize to cytoplasmic bodies also disrupt the ability
of the protein to restrict retroviral infection.
Discussion
In this study, we identify the Linker2, or L2 region of rhTRIM5α as
possessing determinants required for the localization of the protein to
Fig. 3. Quantification of TRIM5α localization and turnover. A. 20 individual Z-stack
images of HeLa cells stably expressing the indicated YFP-rhTRIM5α protein were
collected, deconvolved and analyzed using Imaris image analysis software. The number
of cytoplasmic bodies per cell was determined using fixed size and intensity criteria for
all images. The average number of cytoplasmic bodies identified per cell in all 20 images
is plotted. Error bars represent the SEM. B. HeLa cells stably expressing the indicated
YFP-TRIM5α variant were treated with cyclohexamide for the indicated time period
and YFP-protein expression was analyzed by western blot. Samples were normalized to
include an identical amount of total protein. The graph on the bottom is a densitometric
analysis of the samples shown on the top panel. Results are representative of 3
independent experiments.
Fig. 4. Alanine substitution of residues in the L2 region of rhTRIM5α does not affect
protein multimerization. YFP-WT and the alanine mutants of rhTRIM5α were analyzed
for their ability to form low and higher-order multimers. HeLa cells stably expressing
these proteins were lysed and the cell lysates were treated with increasing
concentration of glutaraldehyde (0, 1, 2, 4 mM) for 5 min at RT. 1 M glycine was
added to saturate the glutaraldehyde, and the protein samples were subjected to SDS-
PAGE. Monomeric and multimeric forms of the protein were identified by immuno-
blotting with a mouse anti-GFP antibody. Results are representative of 3 independent
experiments.
Fig. 5. rhTRIM5α mutants that do not localize to cytoplasmic bodies no longer restrict
HIV-1 infection. HeLa cells stably expressing YFP-WT rhTRIM5α or rhTRIM5α variants
and control HeLa cells were mock infected (DMEM) or infected with serial dilutions of
VSV-g pseudotyped GFP reporter HIV-1. 48 h pi the infected cells were harvested, fixed
and the percentage of GFP positive cells was determined by flow cytometry. HeLa (♦),
WTrhTRIM5α (
), KPK266–268AAA (), RRV275–277AAA (●), TFH269–271AAA
(
), HKN271–273AAA ( ). Results are representative of 3 independent experiments.
263
J. Sastri et al. / Virology 405 (2010) 259–266

Page 6

cytoplasmic bodies. Using scanning alanine mutagenesis, we identify
two stretches of amino acids within this region that are required for
cytoplasmic body localization. Mutations within these regions induce
a diffuse, cytoplasmic localization of protein (Fig. 2) while not
affecting the ability of the protein to dimerize (Fig. 4), demonstrating
that these mutations do not induce a global defect in protein folding.
In all cases tested, the ability to restrict retroviral infection was
abrogated in mutants that exhibited a diffuse, cytoplasmic localiza-
tion. Conversely, mutations that did not disrupt protein localization to
cytoplasmic bodies did not affect the ability to restrict HIV-1 infection.
Whilecytoplasmic body localizationofour epitopetaggedproteins
correlates well with the ability to restrict HIV-1 infection, we do not
believe it is likely that these bodies themselves, as they exist in cells in
the absence of virus, play a role in restriction. Two studies have clearly
demonstrated that pre-existing cytoplasmic bodies are not required
for retroviral restriction (Perez-Caballero et al., 2005; Song et al.,
2005a). Using live cell microscopy, we have previously visualized the
de novo formation of rhTRIM5α cytoplasmic bodies around individ-
ual, fluorescently labeled HIV-1 virions (Campbell et al., 2008). Taken
together with the data presented here, it seems reasonable to
speculate that the residues in the L2 region identified in this study
facilitate the self-association of rhTRIM5α. Studies of the BBox2
domain have found that this domain also mediates TRIM5α self-
association, and that this ability to self-associate is critical to the
ability of TRIM5α to restrict infection (Diaz-Griffero et al., 2009).
Therefore, while the localization of our variants to cytoplasmic bodies
correlates perfectly to the ability to restrict, we favor the hypothesis
that this localization is a reflection of this ability to rapidly form a
multimeric assembly around a virus. Self-associative activities,
mediated by the BBox2 domain and L2 region, likely enhance the
rate at which TRIM5α cytoplasmic assemblies form around restriction
sensitivevirions.However,thedatapresentedherecannotspecifically
exclude the possibility that the pre-existing bodies themselves play
some role in restriction. It is also possible that the L2 region mediates
an interaction with another cellular protein that may be important in
cytoplasmic body formation.
Previousstudies have found that the L2 region can affect the ability
of rhTRIM5α to form low and high-order multimers when analyzed
by biochemical crosslinking (Javanbakht et al., 2005, 2007, 2006;
Mische et al., 2005). Using this technique, we can observe that our L2
variants still form multimers when crosslinked, regardless of their
cellular localization. Glutaraldehyde crosslinking alsoshowedthatour
mutants were able to form higher-order multimers of protein, as
indicated by the higher molecular weight species visible following
treatmentwithincreasingglutaraldehyde concentrations (Fig.4). This
suggests that other domains of TRIM5α, notably the BBox2 domain
(Diaz-Griffero et al., 2009; Li and Sodroski, 2008), may mediate the
formation of higher-order multimers detected using this method.
Other studies examining large deletions of amino acids in the L2
region show that these proteins do not form higher-order multimers,
as detected by glutaraldehyde crosslinking, nor do they form
cytoplasmic bodies (Javanbakht et al., 2007, 2006). Our mutants
retain the ability to form higher-order multimers following glutaral-
dehyde crosslinking while losing their cytoplasmic body localization.
This suggests that cytoplasmic body localization may not be simply
reflective of the ability of the protein to multimerize. It is therefore
tempting to speculate that the L2 region may facilitate the interaction
with a cofactor required for restriction.
The L2 region has not been thoroughly studied, primarily because
this region of the protein is not predicted to form any notable
secondary structure. Because of this fact, it was assumed that this
stretch of amino acids served to functionally tether adjacent domains
of the protein. However, the data presented here suggest a more
important, functional role for this region. This idea is also supported
by the independent evolution of TRIM-Cyp proteins, in which
retrotransposition of the Cyclophilin A gene into the TRIM5α ORF
has functionally replaced the SPRY domain responsible for capsid
binding. It seems noteworthy that all of the transpositions identified
to date, have retained the regions identified in this study as
contributing to cytoplasmic body formation and restriction (Brennan
n et al., 2008; Newman et al., 2008; Sayah et al., 2004; Wilson et al.,
2008).
It is striking that we observe two distinct regions required for
cytoplasmic body localization interceded by a stretch of residues that
are dispensable for cytoplasmic body localization. A similar pattern of
activity was observed in another study in which cyclophilin A was
fused to the TRIM motif of human TRIM5α (Neagu et al., 2009). This
study found that ability of these engineered proteins to restrict HIV-1
depended on the residues in the L2 region to which cyclophilin A was
genetically tethered. These authors did not observe that specific
residues within the L2 region were required, but rather observed that
restriction capacity of their fusion proteins exhibited an apparent
periodicity across the L2 region, with alternating regions of L2
mediating restricting or non-restricting fusion proteins.
Collectively, the data presented here, as well as the works
discussed above, indicate that the Linker2 region of TRIM5α plays a
previously unappreciated role in mediating interactions that are
critical during retroviral restriction. Further studies are required to
determine if this region of TRIM5α mediates the higher-order
multimerization of TRIM5α or is required to mediate an interaction
with other cellular factors involved in the restriction process.
Materials and methods
Recombinant DNA constructs
Wild type rhTRIM5α plasmid was a kind gift from Dr. Joseph
Sodroski (Harvard School of Public Health). HA-tagged rhTRIM5α and
variants were created by inserting SmaI and EcoRI sites flanking
rhTRIM5α using the primers GCCTGGCATTATGCCCAG and AGCTTGC-
CAAACCTAC. This PCR product was then digested with SmaI and EcoRI
and inserted into the EXN retroviral vector, which was generously
provided by the lab of Dr. Greg Towers (Royal Free and University
College, London). The EXN construct was used to derive the YXN
retroviralvector,which wasgeneratedbyPCR amplificationof theYFP
coding region of YFP-N1 (Clontech) using the primers TGGATGAAC-
TATACAAGTGGATCCGGCCG and CGGCCGGATCCACTTGTATAGTT-
CATCCA. This fragment was then digested with AgeI and BsrGI and
inserted into the similarly digested EXN parental plasmid. To facilitate
easier subsequent steps, the BamHI site of wt rhTRIM5α was
disrupted by SOEing PCR using the interior primers CCCCAGTATC-
CAAGCACTTTT and AGTGCTTGGATACTGGGGGTATGT and exterior
primers GCGGCGGGATCCATGGCTTCTGGAATCCT and CGGCCGG-
CTCGAGTCAAGAGCTTGGTGAGC. These primers introduced silent
mutations eliminating the BamHI site present in wt rhTRIM5α. This
PCR product was then inserted into YXN between the BamHI and XhoI
sites of YXN. Alanine mutations were introduced into wt rhTRIM5α or
rhTRIM5α lacking a BamHI site using SOEing PCR. The primers used in
these reactions are provided in Supplementary data table 1.
Cell culture, transfection and virus production
HeLa and 293T cells were cultured in complete DMEM containing
10% fetal bovine serum, penicillin (final concentration 100 U/ml), and
streptomycin (final concentration 100 μg/ml).
HIV reporter virus was produced by Polyethylenimine (PEI)
transfection (Durocher et al., 2002) of 293T cells with 8 μg of pVSV-
G and 12 μg of the proviral construct R7Δ EnvGFP in which the Nef
gene was replaced with GFP. Virus and vector were harvested
identically as previously described (Campbell et al., 2004). To assess
virus infectivity, equivalent numbers of cells in a 24 well plate were
infected for 14 h, after which virus was removed, normal medium
264
J. Sastri et al. / Virology 405 (2010) 259–266

Page 7

added, and GFP expression was determined 48–72 h after infection
using a FACS Canto II flow cytometer (Becton Dickinson).
Stable cell lines
Hela cells were transduced with retroviral vector overnight and
G418 (400 μg/mL) was used to generate cell lines expressing the
indicated, epitope tagged rhTRIM5α variants. The expression of
polyclonal or single colony clones were screened by immunofluores-
cence to ensure all cells expressed the transduced protein. Such cell
lines were then analyzed by western blot analysis, and the clones
expressing comparable amounts of protein were chosen for subse-
quent analysis.
Image analysis
20 Z-stack images of the indicated cell lines were acquired using
identical acquisition parameters. Individual coverslips were coded
such that the individual acquiring the images did not know the
identity of the cell lines. Deconvolved images were analyzed for
cytoplasmic bodies using the Surface Finder function of the Imaris
software package (Bitplane). Surfaces for cytoplasmic bodies in all
samples analyzed were identified using defined fluorescence intensity
and size criteria (Volume=above 0.066 μm3). The number of
cytoplasmic bodies/cell was obtained for each cell line. The data
were plotted in Prism (Graphpad Software Inc) for statistical analysis.
Dunnett's Multiple Comparison test was used to determine the
statistical significance between samples (PU≤U0.05). Unpaired t-Test
with Welch's correction was used to determine the statistical
significance between the KPK266–268AAA and RRV275–277AAA
mutants (pb0.0001).
Western blotting and glutaraldehyde crosslinking assay
Whole cell lysates were prepared by treating 1×105cells with NP-
40 lysis buffer (100 mM Tris pH 8.0, 1% NP-40, 150 mM NaCl)
containing protease inhibitor cocktail (Roche) for 15 min on ice.
Coomassie Plus Bradford Assay (Thermo scientific) was used to
determine total protein concentration. 2× SDS sample buffer was
added to the cell lysates and the samples were boiled for 5 min at
100 °C. Equalamountof protein wasloaded into a 10% polyacrylamide
gel for SDS-polyacrylamide gel electrophoresis (SDS-PAGE). After
separation, the proteins were transferred to nitrocellulose membrane
and detected by incubation with anti-GFP (Covance) or anti-HA
(clone 3F10) conjugated to Horseradish Peroxidase (HRP) (Roche).
Secondary antibodies conjugated to HRP (Thermo Scientific) were
used where necessary and antibody complexes were detected using
SuperSignal™ West Femto Chemilluminescent Substrate (Thermo
Scientific). Chemiluminescence was detected using the UVP EC3™
Imaging System (UVP LLC). Glutaraldehyde crosslinking assays were
performed as previously described (Javanbakht et al., 2005). Briefly,
the cell lysates were incubated on ice for 30 min and centrifuged at
3000 rpm for 1 min to remove the cell debris. The supernatant was
divided into 20 μL aliquots and incubated with 0, 1, 2 and 4 mM
glutaraldehyde for 5 min at room temperature. The glutaraldehyde
was saturated by adding 1 M glycine. 2× SDS sample buffer was added
and the mixture was boiled for 5 min at 100 °C. The samples were
then subjected to SDS-PAGE using 4%–15% Tris–HCl gradient gels
(Ready Gels, BioRad) and subsequent Western Blot analysis.
Protein turnover assay
Cell lines stably expressing the indicated TRIM5α variant were
treated with cyclohexamide (20 μg/ml) and cells were harvested at
the indicated time following cyclohexamide addition. Coomassie Plus
Bradford Assay (Thermo scientific) was used to determine total
protein concentration. Equivalent amounts of protein from individual
samples were subjected to SDS-PAGE and the YFP-TRIM5α protein
was detected by western blot.
Infectivity assay
Equivalent numbers of cells (0.75×105) in a 24-well plate were
infected with VSV-g pseudotyped GFP reporter HIV-1 (R7ΔEnvGFP)
for 14 h, after which the virus was removed and normal DMEM was
added. Percentage of GFP positive cells was determined 48 hpi using a
FACS Canto II flow cytometer (Becton Dickinson).
Acknowledgments
This work was supported by National Institutes of Health grant
K22 AI078757 to EC.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.virol.2010.06.015.
References
Balastik, M., Ferraguti, F., Pires-da Silva, A., Lee, T.H., Alvarez-Bolado, G., Lu, K.P., Gruss,
P.,2008. Deficiency in ubiquitin ligase TRIM2 causes accumulationof neurofilament
light chain and neurodegeneration. Proc. Natl. Acad. Sci. U. S. A. 105 (33),
12016–12021.
Barr, S.D., Smiley, J.R., Bushman, F.D., 2008. The interferon response inhibits HIV particle
production by induction of TRIM22. PLoS Pathog. 4 (2), e1000007.
Bass, S.H., Mulkerrin, M.G., Wells, J.A., 1991. A systematic mutational analysis of
hormone-binding determinants in the human growth hormone receptor. Proc.
Natl. Acad. Sci. U. S. A. 88 (10), 4498–4502.
Brennan, G., Kozyrev, Y., Hu, S.L., 2008. TRIMCyp expression in Old World primates
Macaca nemestrina and Macaca fascicularis. Proc. Natl. Acad. Sci. U. S. A. 105 (9),
3569–3574.
Campbell, E.M., Nunez, R., Hope, T.J., 2004. Disruption of the actin cytoskeleton can
complement the ability of Nef to enhance human immunodeficiency virus type 1
infectivity. J. Virol. 78 (11), 5745–5755.
Campbell, E.M., Dodding, M.P., Yap, M.W., Wu, X., Gallois-Montbrun, S., Malim, M.H.,
Stoye, J.P., Hope, T.J., 2007. TRIM5 alpha cytoplasmic bodies are highly dynamic
structures. Mol. Biol. Cell 18 (6), 2102–2111.
Campbell, E.M., Perez, O., Anderson, J.L., Hope, T.J., 2008. Visualization of a proteasome-
independent intermediate during restriction of HIV-1 by rhesus TRIM5alpha. J. Cell
Biol. 180 (3), 549–561.
Cunningham, B.C., Wells, J.A., 1989. High-resolution epitope mapping of hGH-receptor
interactions by alanine-scanning mutagenesis. Science 244 (4908), 1081–1085.
Diaz-Griffero, F., Vandegraaff, N., Li, Y., McGee-Estrada, K., Stremlau, M., Welikala, S., Si,
Z., Engelman, A., Sodroski, J., 2006. Requirements for capsid-binding and an effector
function in TRIMCyp-mediated restriction of HIV-1. Virology 351 (2), 404–419.
Diaz-Griffero, F., Kar, A., Perron, M., Xiang, S.H., Javanbakht, H., Li, X., Sodroski, J., 2007.
Modulation of retroviral restriction and proteasome inhibitor-resistant turnover by
changes in the TRIM5{alpha} B-box 2 Domain. J. Virol. 81 (19), 10362–10378.
Diaz-Griffero,F.,Qin,X.R.,Hayashi,F.,Kigawa,T.,Finzi,A.,Sarnak,Z.,Lienlaf,M.,Yokoyama,S.,
Sodroski, J., 2009. A B-box 2 surface patch important for TRIM5alpha self-association,
capsid binding avidity, and retrovirus restriction. J. Virol. 83 (20), 10737–10751.
Durocher, Y., Perret, S., Kamen, A., 2002. High-level and high-throughput recombinant
protein production by transient transfection of suspension-growing human 293-
EBNA1 cells. Nucleic Acids Res. 30 (2), E9.
Gack, M.U., Shin, Y.C., Joo, C.H., Urano,T., Liang,C., Sun, L., Takeuchi, O., Akira, S.,Chen, Z.,
Inoue, S., Jung, J.U., 2007. TRIM25 RING-finger E3 ubiquitin ligase is essential for
RIG-I-mediated antiviral activity. Nature 446 (7138), 916–920.
Javanbakht, H., Diaz-Griffero, F., Stremlau, M., Si, Z., Sodroski, J., 2005. The contribution
of RING and B-box 2 domains to retroviral restriction mediated by monkey
TRIM5alpha. J. Biol. Chem. 280 (29), 26933–26940.
Javanbakht, H., Yuan, W., Yeung, D.F., Song, B., Diaz-Griffero, F., Li, Y., Li, X., Stremlau, M.,
Sodroski, J., 2006. Characterization of TRIM5alpha trimerization and its contribu-
tion to human immunodeficiency virus capsid binding. Virology 353 (1), 234–246.
Javanbakht, H., Diaz-Griffero, F., Yuan, W., Yeung, D.F., Li, X., Song, B., Sodroski, J., 2007.
The ability of multimerized cyclophilin A to restrict retrovirus infection. Virology
367 (1), 19–29.
Kar, A.K., Diaz-Griffero, F., Li, Y., Li, X., Sodroski, J., 2008. Biochemical and biophysical charac-
terization of a chimeric TRIM21–TRIM5alpha protein. J. Virol. 82 (23), 11669–11681.
Langelier, C.R.,Sandrin, V.,Eckert, D.M.,Christensen, D.E., Chandrasekaran, V., Alam,S.L.,
Aiken, C., Olsen, J.C., Kar, A.K., Sodroski, J.G., Sundquist, W.I., 2008. Biochemical
characterization of a recombinant TRIM5alpha protein that restricts human
immunodeficiency virus type 1 replication. J. Virol. 82 (23), 11682–11694.
265
J. Sastri et al. / Virology 405 (2010) 259–266

Page 8

Li, X., Sodroski, J., 2008. The TRIM5alpha B-box 2 domain promotes cooperative binding
to the retroviral capsid by mediating higher-order self-association. J. Virol. 82 (23),
11495–11502.
Meroni, G., Diez-Roux, G., 2005. TRIM/RBCC, a novel class of ‘single protein RING finger’
E3 ubiquitin ligases. BioEssays 27 (11), 1147–1157.
Mische, C.C., Javanbakht, H., Song, B., Diaz-Griffero, F., Stremlau, M., Strack, B., Si, Z.,
Sodroski, J., 2005. Retroviral restriction factor TRIM5alpha is a trimer. J. Virol. 79
(22), 14446–14450.
Neagu, M.R., Ziegler, P., Pertel, T., Strambio-De-Castillia, C., Grutter, C., Martinetti, G.,
Mazzucchelli, L., Grutter, M., Manz, M.G., Luban, J., 2009. Potent inhibition of HIV-1
by TRIM5-cyclophilin fusion proteins engineered from human components. J. Clin.
Invest. 119 (10), 3035–3047.
Newman, R.M., Hall, L., Kirmaier, A., Pozzi, L.A., Pery, E., Farzan, M., O'Neil, S.P., Johnson,
W., 2008. Evolution of a TRIM5-CypA splice isoform in old world monkeys. PLoS
Pathog. 4 (2), e1000003.
Ozato, K., Shin, D.M., Chang, T.H., Morse III, H.C., 2008. TRIM family proteins and their
emerging roles in innate immunity. Nat. Rev. Immunol. 8 (11), 849–860.
Perez-Caballero, D., Hatziioannou, T., Zhang, F., Cowan, S., Bieniasz, P.D., 2005. Restric-
tion of human immunodeficiency virus type 1 by TRIM-CypA occurs with rapid
kinetics and independently of cytoplasmic bodies, ubiquitin, and proteasome
activity. J. Virol. 79 (24), 15567–15572.
Reymond, A., Meroni, G., Fantozzi, A., Merla, G., Cairo, S., Luzi, L., Riganelli, D., Zanaria, E.,
Messali, S., Cainarca, S., Guffanti, A., Minucci, S., Pelicci, P.G., Ballabio, A., 2001. The
tripartite motif family identifies cell compartments. EMBO J. 20 (9), 2140–2151.
Sawyer, S.L., Wu, L.I., Emerman, M., Malik, H.S., 2005. Positive selection of primate
TRIM5alpha identifies a critical species-specific retroviral restriction domain. Proc.
Natl. Acad. Sci. U. S. A. 102 (8), 2832–2837.
Sayah, D.M., Sokolskaja, E., Berthoux, L., Luban, J., 2004. Cyclophilin A retrotransposition
into TRIM5 explains owl monkey resistance to HIV-1. Nature 430 (6999), 569–573.
Sebastian, S., Grutter, C., de Castillia, C.S., Pertel, T., Olivari, S., Grutter, M.G., Luban, J.,
2009. An invariant surface patch on the TRIM5alpha PRYSPRY domain is
required for retroviral restriction but dispensable for capsid binding. J. Virol. 83
(7), 3365–3373.
Song, B., Diaz-Griffero, F., Park, D.H., Rogers, T., Stremlau, M., Sodroski, J., 2005a.
TRIM5alpha association with cytoplasmic bodies is not required for antiretroviral
activity. Virology 343 (2), 201–211.
Song, B., Gold, B., O'Huigin, C., Javanbakht, H., Li, X., Stremlau, M., Winkler, C., Dean, M.,
Sodroski, J., 2005b. The B30.2(SPRY) domain of the retroviral restriction factor
TRIM5alpha exhibits lineage-specific length and sequence variation in primates.
J. Virol. 79 (10), 6111–6121.
Stremlau, M., Owens, C.M., Perron, M.J., Kiessling, M., Autissier, P., Sodroski, J., 2004. The
cytoplasmic body component TRIM5alpha restricts HIV-1 infection in Old World
monkeys. Nature 427 (6977), 848–853.
Wilson, S.J., Webb, B.L., Ylinen, L.M., Verschoor, E., Heeney, J.L., Towers, G.J., 2008.
IndependentevolutionofanantiviralTRIMCypinrhesusmacaques.Proc.Natl.Acad.
Sci. U. S. A. 105 (9), 3557–3562.
Yap, M.W., Nisole, S., Stoye, J.P., 2005. A single amino acid change in the SPRY domain of
human Trim5alpha leads to HIV-1 restriction. Curr. Biol. 15 (1), 73–78.
266
J. Sastri et al. / Virology 405 (2010) 259–266