Abstract we will focus on another member

Abstract

HIV invasion in target cells is
hampered by restriction factors that are cell-intrinsic proteins dominantly
acting to potently inhibiting HIV replication. Among these factors, members
of the tripartite motif (TRIM) family have emerged as important players with antiviral
function and in innate immunity. TRIM5a is
the best characterized family member, however, in this review, we will focus on
another member of the TRIM family, i.e. TRIM22. We will discuss TRIM22 as
inhibitor of HIV transcription through a mechanism of interference with the
binding of the transcription factor Sp1 to the HIV promoter. We will then
discuss TRIM22 potential role in HIV-infected individuals debating on its dual
role in controlling HIV replication but also promoting chronic infection.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

 

 

 

1.    
Introduction

Innate
immunity represents the front line against invading pathogens in their host. In
this respect, invasion of viruses in target cells is sensed by a mechanism that
involves both an evolutionary conserved pathogen-associated molecular patterns
(PAMPs) and a germ line encoded pattern recognition receptors (PRRs) {Cao, 2016
#569}. HIV does not evade this rule as early reverse transcriptase products are
detected by at least two intracellular PRRs, IFI16 and cGAS {Altfeld, 2015
#570}. These interactions initiate a signaling cascade that ultimately leads to
transcription of proinflammatory cytokines and type I interferons. The latter binds
to the interferon receptor alpha/beta to induce hundreds of genes with
antiviral function. In addition to the mechanism of PAMPS and PPR recognition,
cells are endowed with intrinsic factors that are expressed prior to infection
and guide the virus to an abortive infection {Malim, 2008 #381}. These factors
are defined altogether as restriction factors (RF). The presence of RF traces
back to 1996 when Friend-virus-susceptibility-1 (Fv1) gene was described to encode a protein able to suppress murine
leukemia virus (MLV) immediately after entry and before integration in the
mouse genome {Best, 1996 #156}. However, the field of RF exploded
in 2002 after the identification of APOBEC proteins {Sheehy, 2002 #488} and,
later in 2004, after the identification of TRIM5 as RFs of HIV and SIV
infection {Stremlau, 2004 #571}. Since then, many other restriction factors
have been identified {Merindol, 2015 #572}. In general, RF are encoded by
individual genes and they dominantly inhibit retrovirus replication in cell
culture {Sheehy, 2002 #488}. However, lentiviruses have evolved accessory viral
proteins that counteract the effect of RF {Towers, 2014 #573}. This feature,
however, is not shared by all RF as some of them, like TRIM5, is not
antagonized by a viral protein {Towers, 2007 #125}. Nevertheless, RF are
characterized by several single nucleotide polymorphisms that represent a
signature of an arms race between pathogen and host during evolution {Sawyer,
2007 #479}. As RF are executors of an innate immunity response, besides their
basal expression, most of them are induced by type I IFN, {Nisole, 2005 #574} a
feature that often endows them with antiviral activity against different virus families.

2.      
TRIM proteins

The
first definition of TRIM family dates back to 2001 {Reymond, 2001 #456} when a
functional genomic study characterized thirty-seven mammalian proteins,
previously clustered into the also known RBCC family {Borden, 1998 #167;Reddy,
1992 #454}, for their common capability to homo-multimerize and to define
specific cell compartments. Since this description, more than seventy TRIM
family members have been identified and, according to the principle that nature
often exploits the same structure to accomplish different functions, TRIM
proteins have been involved in innate immunity, cancer, development and genetic
diseases {Hatakeyama, 2017 #575;Lazzari, 2016 #576;van Tol, 2017 #581}.

The
common and conserved tripartite motif, also known as “RBCC”, forms the
N-terminus of TRIM proteins. This structural motif comprises three functional
domains that are maintained in a precise spatial order throughout evolution {Reymond,
2001 #456}: a RING (R), one or two B-box(es) (B) and a coiled-coil (CC) domain,
suggesting a modular function of the tripartite motif, rather than an
aggregation of distinct units (Figure 1).
Conversely, different domains are located at the C-terminus. Thus far, ten
unrelated domains, which can be present alone or in combination, have been
described. In 2006, Short and Cox proposed a classification of TRIM proteins
according to their composition of sequences in their C-terminal region {Short,
2006 #577}. Thus, eleven subfamilies structurally different in their
C-terminus, are currently listed in the TRIM family (Figure 1). The most frequent motif at the C-terminus of TRIM
proteins is the so-called PRYSPRY or B30.2 that is composed by sixty amino
acids for the PRY domain and the one hundred-and-forty amino acids for the SPRY
domain. The number of TRIM proteins harboring the B30.2 motif is higher than 30
{Sardiello, 2008 #579} and most of the TRIM proteins reported to possess an
antiviral activity are endowed with this domain {Uchil, 2008 #578}.

TRIM22

TRIM22,
also known as “Stimulated Trans-Acting Factor of 50kDa (Staf-50)”, belongs to
C-IV group of TRIM family (Figure 1.12).
Although an increasing number of evidences has highlighted TRIM22 antiviral
activity against several viruses, less is known about biological function of
its domains and its tertiary structure has not been yet resolved. Tripartite
motif of TRIM22 includes a RING domain endowed with E3 Ubiquitin-ligase
activity (Meroni and
Diez-Roux, 2005). In this motif, cysteine-15 and
cysteine-18 constitute the catalytic site that mediates the transfer of
ubiquitin to the target proteins {Eldin, 2009 #243;Lorick, 1999 #580}. The RING
domain of TRIM22 promotes also an auto-ubiquitination pathway that leads to
proteasomal degradation of the protein itself {Duan, 2008 #233;Eldin, 2009
#243}. One B domain only (B-box 2) is present downstream RING domain, likely
required for its nuclear localization (Sivaramakrishnan et al., 2009a), whereas the role of CC domain of TRIM22 is
still unsolved. TRIM22 has also been shown to form trimers (Li et al.,
2007), but it is unknown whether this
high-order self-association plays a function for its biological activity. C
terminus of TRIM22 contains a PRYSPRY domain, also known as B30.2, consisting
in three hypervariable regions. This domain is essential for specific
hetero-dimerization of rhTRIM5? together with HIV-1 capsid protein as mechanism of viral restriction,
however there are no evidences so far of a similar mechanism exploited by
TRIM22 B30.2 (ref).

TRIM22
has been mainly characterized for its antiviral activity against different
viruses. TRIM22 mediated viral restriction of the Picornaviridae encephalomyocarditis virus (EMCV) via interaction of
its B30.2 domain with EMCV 3C protease leading to target polyubiquitination and
repression of viral replication (Eldin et al.,
2009). In addition, TRIM22 inhibited the
replication of the Hepadnaviridae
hepatitis B virus (HBV) by interfering with the HBV core promoter (Gao et al.,
2009). More recently, E3 ubiquitin ligase
activity of TRIM22 has been identified as responsible for the degradation of
the Orthomixoviridae Influenza A
virus (IAV) nucleoprotein, which is essential for viral replication (Di Pietro et al., 2013) and of the Flaviviridae human Hepatitis C NS5A IFN-resistance protein in
response to IFN? (Yang et al.,
2015).

3.1 TRIM22 restriction of HIV infection

            The first report showing that TRIM22
had an anti-HIV activity dates back in 1995 when Nadir Metchi reported for the
first time that TRIM22 inhibits HIV transcription (ref). Later on, Barr
reported that TRIM22 does not inhibit HIV transcription but, through its E3
ubiquitin ligase activity inhibits HIV particle production and release (barr).  In 2011, by studying HIV infection in cell
clones derived from the U937 cells (Franzoso et
al., 1994), we identified TRIM22 selected
expressed in cells non permissive to HIV infection whereas cells permissive to
HIV infection were completely devoid of TRIM22 expression (Kajaste-Rudnitski et al., 2011). By knocking down TRIM22 in non-permissive
cells, virus replication was rescued whereas overexpression of TRIM22 in
permissive cells inhibited HIV replication. The same inhibitory effect was also
reproduced in the A3.01 human T cell line. In search for the molecular
mechanism,  we found that TRIM22
inhibited both basal and phorbol myristate acetate (PMA) plus ionomycin-induced
HIV-1 transcription in 293T cells transfected with a luciferase (Luc) reporter
under the control of HIV-1 LTR (Kajaste-Rudnitski et al., 2011). TRIM22 inhibitory effects were
demonstrated to be independent of NF-kB binding sites, whereas HIV-1 Tat-mediated
transactivation was insensitive to TRIM22 inhibition. Furthermore, a RING
deletion mutant of TRIM22 retained its ability to inhibit HIV-1 transcription,
demonstrating that E3-ubiquitin ligase activity of the protein was not involved
in the mechanism of TRIM22-mediated transcriptional repression of HIV-1. Bu
using infectious molecular clones designed to be independent on tat-TAR
interaction, we then demonstrated that TRIM22 specifically targets HIV
transcription driven by Sp1, that is a cellular transcription factor
indispensable for basal HIV transcription.

It is
well established that basal HIV-1 transcription is mainly driven by the binding
of Sp1 to three consensus binding sites present in the core enhancer of the
HIV-1 LTR upstream of the TATA box (Jones et al.,
1986). These sites are indispensable for
HIV-1 replication as their deletion or mutations suppresses virus replication (Perkins et
al., 1993b; Van Lint et al., 1997a). Sp1 is a 95- to 105-kDa
constitutively expressed transcription factor that activates the promoter of housekeeping genes from positions proximal to
the initiation sites (Rozenberg et al.,
2008). Sp1 binds to the DNA consensus sites
through C-terminal zinc finger motifs (Kadonaga et
al., 1987; Kadonaga et al., 1988) and activates cellular
transcription by interacting with TBP, TATA binding protein-associated factor
(TAF) 110, and RNA polymerase II (Emili et al.,
1994; Xiao et al., 1997). However, how exactly the
interaction with these co-activators regulates the transcriptional activity of
Sp1 has not been completely clarified yet. Although Sp1 was initially
considered as constitutive activator of housekeeping genes, emerging evidences
indicate that post-translational modifications also play an important role in
fine-tuning its function. Phosphorylation of Sp1 is involved in the DNA binding
(Armstrong et al., 1997; Chuang et al., 2012; Fojas de Borja et al., 2001; Tan et al., 2008), which is also influenced by
acetylation (Hung et al.,
2006; Waby et al., 2010), and directly correlates with the
transactivation activity of the protein (Chu, 2012; Courey et al., 1989; Courey and Tjian, 1988; Jackson et al., 1990; Kadonaga et
al., 1988). In addition, glycosylation (Su et al.,
1999), ubiquitination and sumoylation (Gong et al.,
2014; Wang et al., 2011) have been reported to affect Sp1
protein stability. In this regard, although TRIM22 in endowed with an E3
ubiquitin ligase activity, TRIM22 interaction with Sp1 did not result in its
downregulation suggesting that TRIM22 does not promote Sp1 polyubiquitination
and degradation of Sp1 (ref). In addition, TRIM22 did not change Sp1 overall
phosphorylation state that is required for Sp1 transcription activity (ref).

       A number of studies have reported that the Sp1 protein can
serve as an anchor for binding of other factors, promoting a transcriptionally
active environment. As a statement, Sp1 is able to recruit TBP to promoter
lacking of TATA box (Butler and
Kadonaga, 2002). Furthermore, in microglial and
Jurkat T cell lines, Sp1 interacts with chicken ovalbumin upstream promoter
transcription factor (COUP-TF) to synergistically stimulate transcription (Rohr et al.,
1997; Rohr et al., 1999). The ability of Sp1 to enhance
transcription via anchoring different factors is paradigmatically exploited by
HIV-1, whose proviral 5′ LTR is characterized by the juxtaposition of several
transcription factor DNA-binding sites. In particular, the proximity of two
NF-kB and three Sp1 sites (typical of clade B viruses) serves as a platform for
both the transcription factors, which interact and cooperatively promote HIV-1
transcription (Perkins et al., 1994). However, emerging evidences indicate the
existence of cellular proteins directly targeting the Sp1-driven transcription
of HIV-1. COUP-TF-interacting
protein 2 (CTIP2) is able to repress both Tat-driven and earlier phase of HIV-1
transcription by complexing either Tat or Sp1 with heterochromatin protein 1
(HP1) into transcriptionally inactive regions (Marban et al.,
2005b; Rohr et al., 2003). In human microglial cells, CTIP2
promotes the establishment of a heterochromatin environment at the HIV-1
promoter by the recruitment of HDAC 1 and 2 and of the histone
methyltransferase SUV39H1 (Marban et al.,
2007). In addition, a more general role
of CTIP2 as transcriptional repressor has been recently described as a synergism
occurs with the chromatin master regulator High mobility group protein A1
(HMGA1). In particular, HMGA1 is able to recruit CTIP2 and bind the cellular
P-TEFb inhibitor complex 7SK RNA-HEXIM1, thus acting as repressor of both HIV-1
provirus and cellular gene transcription (Eilebrecht et al., 2014). Furthermore, also the
proto-oncogene c-Myc is able to recruit HDAC1 and to form a ternary complex
with Sp1 onto the HIV-1 LTR whereby maintaining the transcriptional repression
state in chronically infected cell line (Jiang et al.,
2007). When HIV-1 transcription is
repressed, HIV-1 provirus persists in a silent state within the genome of
latently infected reservoirs. These findings suggest the hypothesis that TRIM22
might indirectly favor a state of HIV latency.

3.2 TRIM22 as a potential HIV
latency factor

It is
well established that HIV-1 latency is a multifactorial process depending on
epigenetic modification of either chromatin structure or HIV-1 promoter, as
well as on the presence of transcriptional repressors and the availability of
inducible cellular transcription factors (Lusic and
Giacca, 2014; Mbonye and Karn, 2014; Van Lint et al., 2013). Thus, the presence of either CTIP2
or c-Myc in a complex together with Sp1 and HDAC1 suggests that the activity of
these two HIV-1 negative regulators favor a state of HIV-1 latency and can be a
target for therapeutic approaches. In this context, TRIM22 by interfering with
Sp1 might indirectly favor a state of HIV latency.  As TRIM22 target HIV-1 basal transcription
occurring after the proviral integration in the host genome, when low amount of
Tat is encoded. Although TRIM22 possesses a E3-ubiquitin ligase activity and
Sp1 can be ubiquitinated, TRIM22 does not interact directly with Sp1 and,
therefore, does not cause Sp1 downregulation. Therefore, either the stability
or the overall phosphorylation of Sp1 are likely not affected by TRIM22,
suggesting that this post-translational modification is not involved in the
TRIM22-mediated inhibition. However, we cannot exclude that modification at the
level of specific Sp1 amino acidic residues does occur in presence of TRIM22,
as well as the contribution of other Sp1 post-translational modifications in
this transcriptional inhibition. 

As
TRIM22 has an E3 ligase activity, previous reports have focused on viral
targets as substrate for TRIM22-mediated ubiquitination (Barr et al.,
2008a; Di Pietro et al., 2013; Eldin et al., 2009; Yang et al., 2015). Indeed, we have previously shown
that the inhibition of HIV transcription is independent of TRIM22 E3
ubiquitin-ligase activity (Kajaste-Rudnitski et al., 2011) Indeed, in overexpression
conditions or endogenous TRIM22 expression, Sp1 binding to the HIV-1 LTR is
severely impaired, albeit in the absence of a direct Sp1 displacement, as
TRIM22 is not a DNA binding protein and does not bind to the HIV LTR.

DNA
damage caused by HIV proviral integration triggers a DNA-dependent protein
kinase (DNA-PK) that in turns phosphorylates the oncosuppressor protein p53,
inducing the cell death of activated primary CD4+ lymphocytes (Cooper et al.,
2013). This observation suggests that
suppression of the DNA-PK activity in CD4+ cells could facilitate
the formation of latently infected cells, giving rise to the HIV reservoirs in vivo. However, after HIV-1
integration, DNA-PK has a potential dual role as it phosphorylates p53,
inducing an apoptotic cell death pathway, but also Sp1, that activates HIV
basal transcription (Bob Tjean). However, TRIM22 is up-regulated by p53, and
prevents Sp1 binding to the HIV-LTR thus silencing HIV transcription and
favoring HIV latency(Obad et al.,
2004). Of importance is the regulation of
both DNA-PK and TRIM22 in the context of resting vs. activated CD4 T cells and
macrophages. It is known that T cell activation or that can differ depending on
the activation state of the cells and cell type. In U937 promonocytic tumor cell
line, TRIM22 upregulation by p53 exerted an antiproliferative effect inhibiting
the clonogenic growth (Obad et al.,
2004). However, this effect was not
observed in CD3/CD2/CD28-activated CD4+ T lymphocytes suggesting a
more complex role of TRIM22 during T cells activation (Obad et al.,
2007) and highlighting that, beside its
antiviral effect, little is known about TRIM22 expression and activity during
the T lymphocyte cell cycle.

 

3.     Role of TRIM22 in vivo

The
first evidence supporting the function of TRIM22 as suppressor of HIV-1 was
provided by Singh and colleagues in 2011. They assessed that during HIV-1
primary infection, high levels of TRIM22 expression in PBMC were associated
with lower viral plasma load and higher CD4+ T cell counts (Singh et al.,
2011). More recently, the same group has shown
that PBMC from patients with either primary or chronic HIV-1 infection have
significantly higher levels of TRIM22 compared to PBMC from HIV-1-negative
individuals (Yang et al.,
2015). However, these association
studies, which were conducted among cohorts of patients from Centre for the
AIDS Program of Research in South Africa (CAPRISA), were in contrast to another
cohort study that has positively correlated TRIM22 expression with HIV-1 RNA
load at viral set-point (Rotger et al.,
2010). In particular, this latter study
analyzed the transcriptome of CD4+ T cells from HIV-1 infected
individuals, also focusing on genetic variants that potentially affect the
expression of genes involved in the viral control. In this regard, both TRIM5 and TRIM22 genes carry many signatures of positive selection most
likely as a result of interaction with different pathogens (Sawyer et al.,
2007). Recently, 25 amino acid residues
of TRIM22 have been identified as site of non-synonymous SNP and many of them
are located nearby putative functional motifs of the protein (Kelly et al.,
2014). Among the various SNPs, two of
them have been identified to differentially restrict both HIV-1 transcription
and replication {Ghezzi, 2013 #44}. In addition, a TRIM22 haplotype of this 2
SNPs (rs7935564A/G and rs1063303C/G) was found more frequently in advanced
progressors than in individuals with a delayed disease progression. As those
two SNPs are located in the putative region of the TRIM22 coiled-coil domain
and this domain has a role in the TRIM protein multimerization, a direct
consequence of the different restriction activity observed between the two
TRIM22 variant could be ascribed to a different capacity to form high-order
molecular complexes. Indeed, if TRIM22 antiviral activity required homodimers,
one could hypothesize that TRIM22 SNP1-G and SNP2-G would cause conformational
rearrangement in the coil-coiled domain altering the ability of the protein to
recruit other TRIM22 partners. In this regard, we cannot exclude that, at the
nuclear level, TRIM22 interacts with other TRIM proteins, like TRIM19 and/or
TRIM28, promoting a transcriptional adverse environment. Indeed, we have
recently demonstrated that TRIM22 expression colocalizes with TRIM19 in nuclear
bodies {Forlani, 2017 #52} that have been demonstrated to be positioned in
proximity to latent HIV-1 genome in order to inactivate HIV transcription {Lucic,
2016 #583}. Alternatively, conformational changes in the TRIM22 coil-coiled
could change the affinity of the protein for Sp1, determining a decreased
ability to inhibit the HIV-1 basal transcription.

As TRIM22
is endowed with a broad multifunctional anti-viral activity, TRIM22 SNPs were associated
with liver disease caused by HCV and HCV-HIV co-infection. Of note is that
rs1063303 was associated with liver fibrosis but not with sustained virological
response to the treatment with pegylated-interferon alpha plus ribavirin therapy
{Medrano, 2016 #582}. However, an additional SNP (rs7113258A) was identified to
form an aplotype with rs1063303C that was associated with achieving sustained
virological response.   

Concluding remarks

In
conclusion TRIM22 suppresses Sp1-dependent transcription by interacting with
Sp1 and preventing its binding to the HIV-1 LTR. These findings reveal a new
potential mechanism affecting HIV-1 latency and might lead to new
pharmacological approaches aimed at purging the reservoir of
replication-competent latently infected cells. Moreover, as its genetic
diversity affects HIV-1 replication, TRIM22 is a potentially novel determinant
of HIV-1 disease severity. Our results may extend the spectrum of TRIM22
antiviral activity to other viruses that exploit the same transcriptional
mechanism and bear relevance to cellular pathway(s) under the control of the
Sp1-dependent transcription.