Modulation of LINE-1 and Alu/SVA Retrotransposition by Aicardi-Goutières Syndrome-Related SAMHD1

1LINE-1 is the only active autonomous retroelement in humans and can produce new genomic insertions mediated by its encoded endonuclease and reverse transcriptase activities. LINE-1s make up at least 17% of the human genome. Other nonautonomous retroelements, such as Alu and SVA, complete their retrotransposition process by a mechanism mediated by LINE-1 proteins and account for approximately 1 million and 3,000 copies, respectively. Recent data have suggested that the activity of retroelements such as LINE-1, Alu, and SVA can lead to various diseases.

2SAMHD1 mutations can cause Aicardi-Goutières syndrome (AGS), which is characterized as an improper immune activation resulting from the accumulation of intracellular DNA. SAMHD1 is also a potent cellular restriction factor against retroviruses such as HIV and simian immunodeficiency virus (SIV), with a deoxynucleoside triphosphate triphosphohydrolase (dNTPase) activity linked to retroviral restriction. SAMHD1 inhibits retroviruses in nondividing myeloid cells and resting CD4+ T cells by depleting dNTP levels, although recent studies suggested that the phosphorylation status of SAMHD1 (T592) is also important for the anti-HIV activity. However, the ability of SAMHD1 to inhibit endogenous retroelements such as LINE-1 has not been previously reported.

3To determine whether LINE-1 is a potential target of SAMHD1, we evaluated the effect of SAMHD1 on LINE-1 retrotransposition using a well-established reporter system in HEK293T cells (Figure S1A). The LINE-1 construct 99 PUR RPS EGFP contains an EGFP reporter cassette, interrupted by an intron in the opposite transcriptional orientation and inserted into the 3′ UTR of a retrotransposition-competent L1, L1-RP (Figure S1B). EGFP is expressed only when the LINE-1 transcript is spliced and reverse transcribed, its complementary DNA (cDNA) is inserted into the host genome, and the EGFP reporter gene is expressed from its own CMV promoter. Construct 99 PUR JM111 EGFP (JM111), which contains two missense mutations in ORF1, produces no EGFP signal (i.e., it is retrotransposition defective) and was used as a negative control for retrotransposition (Figure S1C). 99 PUR RPS EGFP generated approximately 0.8% GFP-positive cells, whereas JM111 produced no EGFP-positive cells (Figure 1A). The empty vector (VR1012) had no apparent effect on LINE-1 activity when compared to a no-vector control (Figure S1D). SAMHD1 potently inhibited LINE-1 retrotransposition in a dose-dependent fashion in HEK293T cells (Figure 1A). The expression of mCherry had no effect on LINE-1 retrotransposition, whereas fusion protein mCherry-SAMHD1 was still potent in LINE-1 inhibition (Figures S1E and S1F). SAMHD1 expression had no apparent effect on cellular proliferation or cellular β-actin mRNA splicing (Figures S1G and S1H). The LINE-1 assay relies on CMV promoter-driven EGFP expression as a readout. Under a similar percentage of EGFP-positive cells, coexpression of SAMHD1 in a dose manner with a CMV-EGFP-expressing vector did not alter the expression of EGFP (Figure 1B), showing that SAMHD1 does not affect either CMV promoter-mediated transcription or translation/stability of the EGFP protein.

4A synthetic human LINE-1 construct (ORFeus-HS) containing codon-modified ORF1 and ORF2 sequences and a deleted 5′ UTR was also suppressed by SAMHD1 (Figures S2A and S2B). Because ORFeus-HS contains little authentic LINE-1 RNA sequence, it appears that SAMHD1 may not target the cis-sequence of LINE-1. Furthermore, retrotransposition of Neo-based L1-RP in HeLa cells was efficiently inhibited by SAMHD1 (Figures S2A, S2C, and S2D). Thus, we have demonstrated that SAMHD1 inhibits retrotrans-position in diverse mammalian LINE-1 systems.

5Endogenous SAMHD1 can also function as a LINE-1 inhibitor. Addition of specific small interfering RNAs (siRNAs) targeting SAMHD1 resulted in a 230% increase in LINE-1 retrotransposition in HEK293T cells (Figure 1C). Similar L1 inhibition was observed when the expression of endogenous SAMHD1 was reduced by using a SIV Vpx (viral protein X) expression vector, but not by a mutant VpxQ76A expression vector (defective in mediating SAMHD1 degradation; ) in HEK293T cells (Figures S2E and S2F).

6Although we achieved approximately 90% transfection efficiency as indicated by the efficient degradation of endogenous SAMHD1 by the transfected Vpx expression vector (Figure S2E) and the detection of mCherry expression after transfection in parallel experiments (data not shown), SAMHD1 expression in HEK293T cells had no obvious effect on the infection of wild-type SIVsmm (SIV-WT) or SIVsmmΔVpx (Figure 1D). It is unlikely that SIV-WT or SIVsmmΔVpx preferentially infected untransfected HEK293T cells, because SIV-WT infection still resulted in transfected SAMHD1 degradation in these cells (Figure 1E). SAMHD1 expression also did not inhibit HIV-1 infection in HEK293T cells (Figure S2G). SAMHD1 restricts the retroviral reverse transcription in nondividing myeloid cells by depleting the intracellular dNTP pool. This dNTP depletion activity is countered by the active production of dNTPs in the dividing cells. Our data showing that SAMHD1 cannot suppress HIV/SIV infection in HEK293T cells are consistent with this model. We also observed that SAMHD1 did not inhibit HBV replication (which also requires dNTPs during reverse transcription) in HEK293T cells (Figure S2H). Therefore, SAMHD1 may inhibit LINE-1 and retroviruses through apparently distinct mechanisms. Consistent with this idea, the SAMHD1 mutant D311A, which is defective for dNTPase activity, maintained the ability to inhibit LINE-1 retro-transposition (Figure 1F). Active site mutants of SAMHD1 have been reported to lack anti-HIV-1 or SIVΔVpx activity. These results suggest that dNTPase activity may not be critical for SAMHD1-mediated LINE-1 inhibition. We have also observed that phosphorylation status of SAMHD1 (T592), which is important for HIV-1 restriction, did not affect LINE-1 inhibition (Figure 1G).

7Specific SAMHD1 point mutations, internal deletions, and carboxyl-terminal truncations (Figure 2A) can cause AGS. An examination of several SAMHD1 mutants that have been identified in AGS patients indicated that they all show significantly reduced LINE-1 inhibition (p < 0.01 versus wild-type SAMHD1), even when expressed at levels comparable to wild-type SAMHD1 (Figures 2B and 2C). LINE-1 reverse transcription occurs on genomic DNA in the nucleus through a process called target-primed reverse transcription. SAMHD1 has also been localized to the nucleus. Thus, nuclear localization could be required for SAMHD1-mediated LINE-1 inhibition. However, we found that the SAMHD1 point mutants and SAMHD1 truncation mutants maintained a nuclear localization in live HEK293T cells (Figure S3A; data not shown).

8We further examined regions in SAMHD1 that are important for its anti-LINE-1 activity (Figure 2A). We observed that SAMHD1 mutants lacking the partial HD domain (SAMHD1 Δ162-335) and amino acids 113–136 (SAMHD1 Δ113-136) had weaker activity against LINE-1 than did full-length SAMHD1 (Figures S3B and S3C). On the other hand, deletion of the SAM domain (SAMHD1 Δ42-109 had an increased effect on LINE-1 inhibition (Figures S3B and S3C). However, deletion of the N-terminal region including the linker region abolished SAMHD1-mediated LINE-1 inhibition (Figure 2D). It is worth noting that deletion of SAMHD1 SAM domain reduced its dNTPase and anti-HIV-1 activities. However, SAM domain deletion SAMHD1 mutant still inhibited LINE-1 (Figures S3B and S3C), consistent with the argument that SAMHD1 inhibits LINE-1 and HIV-1/SIV through distinct mechanisms. Thus, in SAMHD1, the linker region between the SAM and HD domains, and possibly the HD domain itself, is critical for LINE1 inhibition. Interestingly, the AGS-related SAMHD1 point mutants that have a reduced capacity for LINE-1 inhibition are all clustered in these regions.

9We were also interested in the LINE-1 inhibition potency of SAMHD1 from other animals. SAMHD1 from the nonhuman primates Nomascus leucogenys (NL) and Macaca mulatta (MM) showed strong LINE-1 inhibition when compared to human SAMHD1 (Figure 3A). SAMHD1 from the primate Pongo abelii (PA) had a slightly weaker anti-LINE-1 activity than that of human SAMHD1 (Figure 3A). Also, SAMHD1 from both Canis lupus familiaris (CLF) and Bos taurus (BT) had strong anti-LINE-1 activity (Figure 3B). Interestingly, human and Mus musculus (Mus) SAMHD1 could potently suppress the retrotransposition of LINE-1 from both human and mouse (Figures 3C and 3D). Thus, LINE-1 inhibition appears to be a conserved feature of mammalian SAMHD1 proteins, and SAMHD1 protein functions through a general mechanism to suppress LINE-1 from different species.

10Several attempts were made to identify the potential target of SAMHD1 in LINE-1 retrotransposition. We first tested the effect of SAMHD1 on LINE-1 gene expression. A few transcription factors that interact with LINE-1s have been identified. HnRNPL and RNA helicase MOV10 have been reported to influence LINE-1 ORF1 expression. We have confirmed in our assay system that HnRNPL and MOV10 could reduce LINE-1 ORF1p expression (Figure 4A). Unlike HnRNPL and MOV10, SAMHD1 did not alter LINE-1 ORF1p expression (Figure 4A). LINE-1 replication requires reverse transcription of its own RNA genome using ORF2p. An in vitro LEAP reverse transcriptase assay has been developed to assess the function of ORF2p in LINE-1 ribonucleoprotein (RNP) complexes (Figure 4B). LINE-1 RNPs from an ORF1-tagged LINE-1 construct were isolated from transfected HEK293T cells in the absence or presence of SAMHD1. SAMHD1 did not affect the isolation of LINE-1 RNPs, as indicated by the detection of ORF1 protein or LINE-1 RNA (Figure 4C). However, ORF2p-mediated endogenous reverse transcription of LINE-1 RNA (LEAP products) was significantly suppressed in the presence of SAMHD1 (Figure 4C, lane 2) when compared to its absence (lane 1). Quantitative real-time PCR indicated that SAMHD1 caused an 83% reduction of the ORF2-mediated endogenous reverse transcription of LINE-1 RNA (Figure 4D). Further investigation indicated that SAMHD1 reduced the expression of ORF2p, by 62% in average (Figures 4E and S4A). Moreover, SAMHD1 inhibited ORF2p-mediated Alu and SINE-VNTR-Alu (SVA) retrotransposition activity (Figures S4B–S4E).

11This study has identified LINE-1, Alu, and SVA retroelements as the potential targets of SAMHD1. We have determined that human SAMHD1 inhibits LINE-1 activity using various established human and mouse LINE-1 retrotransposition assay systems. SAMHD1 can also inhibit the retrotransposition of a modified human synthetic LINE-1 that contains little authentic LINE-1 sequence. Thus, SAMHD1 may not target the LINE-1 cis-RNA sequence. Consistent with this argument, we observed no detectable interaction of SAMHD1 with endogenous LINE-1 RNA, although LINE-1 RNA binding with ORF1 protein was readily detected (data not shown).

12The effect of SAMHD1 on LINE-1 activity appeared to be different from other cellular regulators such as HnRNPL and MOV10. HnRNPL and MOV10 have been observed to affect LINE-1 ORF1 expression. On the other hand, we have observed that SAMHD1 did not alter LINE-1 ORF1p expression (Figure 4A). However, SAMHD1 reduced ORF2p expression and suppressed ORF2p-mediated LINE-1 reverse transcription in purified LINE-1 RNP. Furthermore, SAMHD1 also inhibited ORF2p-mediated Alu and SVA retrotransposition. Thus, LINE-1 ORF2p may be a potential target of SAMHD1.

13Interestingly, it seems that SAMHD1 may inhibit retroviruses and retrotransposons through different mechanisms. SAMHD1 inhibits HIV/SIV in nondividing but not dividing cells. In contrast, LINE-1 inhibition by SAMHD1 is observed in dividing cells. SAMHD1’s dNTPase activity depends on residues H167, H206, D207, and D311 within the HD domain and is crucial for retroviral inhibition because mutations at H206/D207 compromise both enzymatic activity and antiviral potency. However, in our study, the D311A mutation, which depletes the hydrolase activity of SAMHD1, still functions as a LINE-1 inhibitor. Elucidating such a mechanism would provide additional critical information to increase our understanding of the process of retrotransposition, as well as providing a potential explanation for SAMHD1 mutations leading to AGS.

14Thus far, five genes (TREX1, RNASEH2A, RNASEH2B, RNASEH2C, and SAMHD1) have been linked to AGS and are related to an inability to remove cellular nucleotide debris intracellularly. There is no direct link between LINE-1 activity and the cellular immune response; however, it has recently been suggested that LINE-1 activity is linked to both interferon expression and selected autoimmune disorders. A recent report concluded that LINE-1 activity is upregulated in TREX1 knockout cells, suggesting that TREX1 is a LINE-1 inhibitor. Here, we provide evidence that another AGS-related protein, SAMHD1, is also a potent LINE-1 suppressor. AGS-related mutations compromise the potency of SAMHD1 against LINE-1 retrotransposition. AGS-related mutations also compromised TREX1’s potency against LINE-1. Interestingly, reverse transcriptase inhibitors could reduce disease symptoms generated in TREX1-knockout mice. Further investigation is required to understand the relationship between retroelements such as LINE-1, AGS-suspected genes, and AGS itself.

15HEK293T cells and HeLa-HA cells were grown in DMEM medium with 10% FBS (HyClone), GlutaMax, and Penstrep (Invitrogen). All transfections used Lipofectamine 2000 (Invitrogen) reagent. HeLa-HA cells were a gift from J.V. Moran (University of Michigan).

16The human L1 plasmids 99 PUR L1RP EGFP (L1) and 99 PUR JM111 EGFP (JM111, as a negative control) have been previously described. Similar protocols were also applied to either the synthetic human LINE-1 sL1-ORFeus-HS or the murine LINE-1 plasmid ORFeus (gifts of Dr. J.D. Boeke). In brief, LINE-1 plasmid was transfected into HEK293T cells at 2 μg per well in 12-well plates, together with VR1012 or one of the test plasmids. The cells were selected by the addition of puromycin (final concentration, 5 μg/ml) at 48 hr posttransfection. GFP-positive cells were examined 48 hr later by flow cytometry using FACSCalibur. Gating exclusions were based on background fluorescence of the plasmid 99 PUR JM111 EGFP, an L1 construct containing two point mutations in ORF1 that completely abolish retrotransposition; 10,000 single-cell events per sample were gated and analyzed using CellQuest Pro (v.5.2). For the mneoI-based LINE-1 retrotransposition assay, 1 μg LcRPS-mneoI was cotransfected in 6-well plates with 0.5 μg empty vector or the SAMHD1 expression vector into HEK293T cells. At 4 days posttransfection, selection with G418 was initiated and continued for 13 days. The cells in T75 flasks were fixed with PBS/paraformaldehyde/glutaraldehyde, and colonies stained with 0.4% Giemsa.

17The LINE-1 construct containing FLAG-HA-tagged ORF1, pc-L1-1FH, has been described. It was cotransfected in the absence or presence of the SAMHD1 expression vector into HEK293T cells. At 2 days after transfection, LINE-1 RNPs were isolated by ultracentrifugation through a sucrose cushion as previously described. The LINE-1 RNP sample (2 μl) was added to each cDNA extension reaction (LEAP) as previously described, using the 3′ RACE adaptor NV: 5′-GCGAGCACAGAATTAATACGACTCACTATAGG TTTTTTTTTTTTVN-3′ as primer. Also, LINE-1 RNA was extracted from the LINE-1 RNP, treated with TURBO DNase (Invitrogen), and reverse transcribed using 3′ RACE adaptor NV as primer and MuLV RT with a High Capacity cDNA Reverse Transcription Kit (ABI Applied Biosystems). PCR was performed as previously described. PCR products were separated on 2% agarose gels and visualized by the Red Personal Gel Imaging System (ProteinSimple). PCR products were also sequenced and found to match the expected LINE-1 sequence. The relative amount of synthesized cDNA from both methods were then determined by real-time PCR using the primers Linker (as part of primer 3′ RACE adaptor NV), 5′-GCGAGCACAGAATTAATACGACT-3′; L1-LEAP-R, 5′-GGGTTCGAAATCGATAAGCTTGGATCCAGAC-3′, with a standard two step method (95°C for 15 s and 60°C for 1 min) with a cycle number of 40. See also Extended Experimental Procedures.