Activation of LTRs from Different Human Endogenous Retrovirus (HERV) Families by the HTLV-1 Tax Protein and T-Cell Activators

Human endogenous retroviruses (HERVs) represent approximately 8% of our genome. HERVs influence cellular gene expression and contribute to normal physiological processes such as cellular differentiation and morphogenesis. HERVs have also been associated with certain pathological conditions, including cancer and neurodegenerative diseases. As HTLV-1 causes adult T-cell leukemia and HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP) and has been shown to modulate host gene expression mainly through the expression of the powerful Tax transactivator, herein we were interested in looking at the potential modulation capacity of HTLV-1 Tax on HERV expression. In order to evaluate the promoter activity of different HERV LTRs, pHERV-LTR-luc constructs were co-transfected in Jurkat T-cells with a Tax expression vector. Tax expression potently increased the LTR activity of HERV-W8 and HERV-H (MC16). In parallel, Jurkat cells were also stimulated with different T-cell-activating agents and HERV LTRs were observed to respond to different combination of Forskolin, bpV[pic] a protein tyrosine phosphatase inhibitor, and PMA. Transfection of expression vectors for different Tax mutants in Jurkat cells showed that several transcription factors including CREB appeared to be important for HERV-W8 LTR activation. Deletion mutants were derived from the HERV-W8 LTR and the region from −137 to −123 was found to be important for LTR response following Tax expression in Jurkat cells, while a different region was shown to be required in cells treated with activators. Our results thus demonstrated that HTLV-1 Tax activates several HERV LTRs. This raises the possibility that upregulated HERV expression could be involved in diseases associated with HTLV-1 infection.


Introduction
An estimated 8% of our genome is derived from Human Endogenous Retrovirus (HERV), sequences which are resulting from integration events that took place millions of years ago. HERVs are known to have endogenized from ancestral exogenous retroviruses during primate evolution. HERVs are classified into 3 classes (class I, II, and III) based on their sequence similarities to different infectious retroviruses. Each class is divided in subgroups, based on the specificity of the tRNA primer-binding site (PBS) [1], i.e., HERV-W uses the tryptophan (W) tRNA as its primer whereas HERV-K uses the lysine (K) tRNA. While most HERVs are defective and unable to produce infectious particles, some of them have retained the capacity to encode viral proteins [2]. In addition, most HERVs have lost their entire coding sequences by homologous recombination between the two LTRs, leaving solitary LTRs [3]. These LTRs remain active in their promoter and can thereby modify the expression of adjacent cellular genes.
HERV genes play an important role in many physiological events such as placental development [4,5], in which HERV-derived syncytin-1 and syncytin-2 genes seem to be the two major players by promoting the cellular fusion of trophoblasts [4,6]. On the other hand, HERVs have also been associated with several human autoimmune diseases and cancer. For instance, evidences suggest that syncytin-1 is involved in breast cancer [7,8] and multiple sclerosis [9]. In addition, a reduction in the expression of the capsid protein of HERV-W was observed in neurons and glial cells from brains of patients with schizophrenia, bipolar disorder and major depression [10], while HERV-W transcripts were more abundant in cerebrospinal fluid and plasma from patients with schizophrenia [11].
Human T-cell lymphotropic virus type-1 (HTLV-1) is the causative agents of adult T-cell leukemia (ATL) and has also been associated with a chronically progressive neuro-inflammatory disease known as HTLV-1 associated myelopathy or tropical spastic paraparesis (HAM/TSP). However, the majority of HTLV-1-infected patients remain asymptomatic throughout their lifetime. The HTLV-1 Tax protein is a powerful transactivator strongly suggested to be determinant in the development of ATL as well as HAM/TSP [12,13]. These links likely result from the capacity of this protein to activate several transcription factors such as CREB, NF-B and SRF, which leads to upregulation or downregulation of a number of cellular genes [14][15][16][17][18].
Previous studies have shown that viruses such as HSV-1 and the Influenza virus could modulate HERV LTR activity [19][20][21]. Given that HERV overexpression has been associated with multiple sclerosis, a disease resembling HAM/TSP, we thereby tested whether the Tax protein could modulate HERV gene expression. Our results indeed confirm, that alike T cell activators, Tax significantly, yet selectively, induced LTR activity of several HERV family representatives.

Different HERV LTRs Are Activated upon T Cell Activation
We first tested whether the activation of T lymphocytes could modulate the expression of HERVs. Different T-cell-activating agents known to activate many transcription factors in T-cells were thus first tested. Jurkat cells were transfected with luciferase reporter vectors harboring 5'LTR from different HERV families. LTRs from HERV-W4, HERV-W8, HERV W18, HERV-H (MC16), HERV-K (TD47) and HERV-E (E2) were thus tested individually for their responsiveness to T-cell activators. As shown in Figure 1A, HERV-W8 and HERV-H (MC16) LTRs were highly responsive to a combination of Forskolin and bpV[pic], a cAMP pathway activator and an inhibitor of protein tyrosine phosphatases, respectively and to the bpV[pic]/PMA combination. Both LTRs were also significantly responsive to the addition of bpV[pic] alone. While HERV-E and HERV-K representative LTRs were not activated by any tested agents, HERV-W4 and HERV-W8 presented a significant induction of LTR activity only in the presence of the Forskolin/bpV[pic] combination, although the response was more modest. To confirm these results, RNA from Jurkat cells was analyzed by RT-PCR for transcript levels of gag or pol genes from HERV-H, HERV-K, HERV-W and HERV-E families ( Figure 1B). In bpV[pic]/Forskolin-stimulated cells, we confirmed that activation indeed led to an increase in HERV-W gag and HERV-H pol transcript levels when compared to untreated cells. A limited modulation of HERV-K gag and HERV-E pol expression was noted upon stimulation, again confirming the results obtained with the LTR constructs. The induction mediated by the bpV[pic]/PMA combination was also specific to HERV-H and HERV-W LTRs and was again less pronounced than the one observed with the Forskolin/bpV[pic] combination.
We and others have previously indicated that the PTP inhibitor bpV [pic] in T-cells can activate a multitude of transcription factors, such as NF-B, NFAT, STAT, AP-1 and CREB [22][23][24][25]. Our results thereby first indicated that activation of T cells and activation of some of these transcription factors led to induction of HERV gene expression. Interestingly, we have previously demonstrated that Forskolin/bpV[pic] strongly induced the LTR of syncytin-1 (HERV-W) and syncytin-2 (HERV-FRD) genes in the choriocarcinoma BeWo cell line [4]. Other studies in T lymphocytes have indicated that HERV-H transcripts could be induced in T-cell leukemia cell lines by PHA [26] Our studies in the T-cell context are thus in line with the induction potential of HERV LTRs [27].

HERV-W
To identi generated fr hese deletio activators. A a combinatio with the  wild-type Ta Figure 4C) observed wi nonetheless our tested T-  [19,20]. More specific studies focused on a HERV-W representative encoding for syncytin-1 have revealed that several transcription factors were acting on basal and cAMP-mediated LTR activation in trophoblast cells, such as GCMa, Sp1, GATA transcription factors and other potential transcription factors [5,[32][33][34]. A potential NF-B-binding site has also been identified for its importance in TNF--mediated activation of syncytin-1 expression in astrocytes [35].
Our results suggest that different transcription factors act upon the induction of HERV-W8 LTR by T-cell activators and Tax. We are currently conducting experiments to more precisely identify these LTR regions.

Generation of Deletions Mutants by Exonuclease III
The pHERV-W8 LTR-Luc construct was first digested with BstX1 and BamHI and subsequently incubated in the presence of exonuclease III at 37 °C. At 15 s intervals, an aliquot was taken and added to a tube containing S1 nuclease. At the end of the time course, the S1 nuclease reaction was completed at 30 °C. Each samples were heat inactivated and after religation, transformed in DH5. Sequencing of plasmid DNA from resulting colonies was conducted for positioning the resulting 5' end of the LTR. Five deletion mutants were chosen for subsequent experiments.

Statistical Analyses
All experiments were performed in triplicates. Results are expressed as the mean+SEM and statistically analyzed using a 2-tailed Student t test for 2-group comparisons.

Conclusions
We herein have demonstrated that LTRs from different HERV families can both be activated by the combination of T-cell activation agent bpV[pic]/Forskolin and bpV[pic]/PMA and by the Tax transactivator. Furthermore, we identified a Tax-responsive region different from the region responsive to T-cell activators suggesting the implication of different transcription factors. Indeed the use of the various Tax mutants reveals that each of them affects this LTR activation at different degrees. Through the CREB dominant negative mutant, our results argue that members of the CREB/ATF family are playing a role in the upregulation of the LTR activity. The HTLV-1 Tax protein is a powerful transactivator, capable of inducing many cellular genes with its activation domains. It is thereby not surprising that it also acts on the HERV-W LTRs, which are known to be upregulated by inducing agents such as Forskolin and bpV[pic] in human trophoblasts and by HSV-1 and Influenza virus infection [4,20,21]. We conclude that Tax can modulate these LTRs in Jurkat cells by activating CREB and possibly NF-қB, which can positively regulate the transcription of the HERVs genes or other cellular genes in proximity. Transcription factors mediating the upregulation of HERV LTRs by both Tax and T-cell activators are likely different and we are currently working on their identification.
Since our results indicate that Tax can modulate the expression of HERVs, a link between HTLV-1-associated diseases and HERV dysregulation is an interesting speculation. Indeed, upregulation of HERV gene expression has been associated with various inflammatory and autoimmune diseases such as multiple sclerosis and arthritis [9,41]. Interestingly, HTLV-1-associated diseases HAM/TSP and HTLV-1-associated arthropathy have been shown to be very similar to these latter diseases. In addition, activation of HERV LTR nearby proto-oncogenes may constitute a mechanism by which Tax could promote cell transformation via HERV sequences. Alternatively, induced expression of near full-length HERV proviral DNA could generate potential substrates for reverse transcriptase activity. Newly synthesized proviruses could then reintegrate the host genome in infected cells and contribute to genomic instability.
More advanced studies are needed to clearly determine if HERV expression is increased in HTLV-1-infected patients and modulated during the course of HTLV-1-induced pathologies. Furthermore, in this study, we have focused on the modulation of HERV-LTR by the viral protein Tax of HTLV-1. Clearly, other viral proteins such as HBZ could impact on the extent of Tax-mediated HERV LTR activation or could affect HERV LTR activation. These experiments, as well as the analysis of HERV expression in HTLV-I-infected cells are currently ongoing. These studies will indicate whether HERVs could become possible new disease markers in HTLV-1-infected patients.