Olivier Ferrigno ¹, Thierry Virolle ¹, Zied Djabari ¹, Jean-Paul Ortonne ¹, Robert J. White ², and Daniel Aberdam ¹

¹ U385 INSERM, Faculté de Médecine, Nice, France.
² Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow, UK.

Correspondence should be addressed to:
D Aberdam.    e-mail: aberdam@unice.fr

Short interspersed elements (SINEs) are highly abundant components of mammalian genomes that are propagated by retrotransposition. SINEs are recognized as a causal agent of human disease and must also have had a profound influence in shaping eukaryotic genomes [1]. The B2 SINE family constitutes approximately 0.7% of total mouse genomic DNA (ref. 2) and is also found at low abundance in humans [3]. It resembles the Alu family in several respects, such as its mechanism of propagation. B2 SINEs are derived from tRNA and are transcribed by RNA polymerase (pol) III to generate short transcripts that are not translated [4, 5]. We find here, however, that one B2 SINE also carries an active pol II promoter located outside the tRNA region. Indeed, a B2 element is responsible for the production of a mouse Lama3 transcript. The B2 pol II promoters can be bound and stimulated by the transcription factor USF (for upstream stimulatory factor), as shown by transient transfection experiments. Moreover, this pol II activity does not preclude the pol III transcription necessary for retrotransposition [6]. Dispersal of B2 SINEs by retrotransposition may therefore have provided numerous opportunities for creating regulated pol II transcription at novel genomic sites. This mechanism may have allowed the evolution of new transcription units and new genes.

 Figure 1: Identification of a pol II promoter in a B2 SINE.

a, The B2 SINE within mouse Lama3 is flanked by 20-bp direct repeats (hatched boxes) and contains pol III promoter elements (A- and B-blocks). The region of B2 sequence with homology to tRNA is indicated, as are the start site (+1), the TATA box of the pol II promoter and the putative ATG translational start site of the Lama3 transcript [7]. The asterisk represents the location of the 5' end of the a3C cDNA previously misassigned by primer extension [7]. Also shown are the orientations of the pol II and pol III transcripts and the restriction sites for AflIII (A), PvuII (P), DraI (D), BstY (BstY), BlpI (B) and EagI (E).

b, In vitro transcription assays using HeLa whole cell extract (lanes 1–3) or a reconstituted system containing purified pol II, TFIIA and TFIIH along with recombinant TBP, TFIIB and TFIIE (lanes 4–7). Templates were 50 ng (lane 1) or 100 ng (lanes 2, 3) of PCR-amplified fragment containing DraI-BlpI (DB), or 50 ng of linearized pGL2 plasmid containing the AflII-EagI (AE, lane 4), DraI-EagI (DE, lane 5), AflII-BlpI (AB, lane 6) or DraI-BlpI (DB, lane 7) restriction fragments. Reaction 3 contained a-amanitin. Run-off transcripts are shown. The precise start site was mapped by alignment with sequencing reactions.

c, RT–PCR assays using poly(A)⁺ RNA extracted from NIH-3T3 cells and mouse lung were performed with alternative nested primers to demonstrate that a Lama3 transcript initiates within the B2 element in vivo. The position and orientation of the primers are indicated by arrows under the AE sequence. Each primer (bold) is tested by PCR using AE construct as template. Amplification products are obtained using primers L3–L5, but not with primers L1 or L2; this places the in vivo start site between primers L2 and L3. No amplification is detected when the reverse transcriptase was omitted (not shown). The transcription start site mapped in (b) is boxed (+1).

d, The Lama3 B2 SINE supports expression of the luciferase gene in vivo. NIH-3T3 cells were transfected with luciferase reporter pGL2-basic driven by the different Lama3 B2 fragments, TK or VEGF gene minimal promoters. Values shown are given relative to the TK construct, which is assigned a value of 100.

Figure 2: Characterization of the minimal pol II promoter of the Lama3 B2 SINE.

a, Sequence comparison of the initiator regions of the Lama3 B2, adenovirus IV_a2, terminal deoxynucleotidyltransferase (TdT) and adenovirus major late (MLP) promoters. The transcription initiation site is designated +1.

b, The pol II promoter of Lama3 B2 contains TATA box and initiator elements that function in vivo. NIH-3T3 cells were transfected with CMV-_bgal and a luciferase reporter subcloned downstream of the wild-type or substituted DB or DBstY fragments, as indicated. Values are given relative to the DB construct, which was assigned a value of 100. Bottom, sequence changes introduced into the TATA box and initiator regions of the DB fragment of the Lama3 B2 SINE.

The key elements of many pol II core promoters are a TATA box and an initiator element [9]. Pol II transcription from Lama3 B2 was found to begin at a site resembling the initiator consensus sequence (Fig. 2a). Moreover, a TATA box homology lies 25 bp upstream of this start site within an A/T-rich region (Fig. 2b). To test whether these putative elements are responsible for the observed pol II transcription, we mutated them in the context of the 160-bp DB fragment (Fig. 2b). Substitution of 7 bp encompassing the putative TATA box reduced expression of the luciferase reporter by 75% in vivo (Fig. 2b, DB/TATAmut). Moreover, a 10-bp substitution encompassing the initiator homology abolished luciferase production (Fig. 2b, DB/inrmut). In contrast, deleting 90 bp from downstream of the initiator region had only a minor effect (Fig. 2b, DBstY). We therefore conclude that pol II transcription in vivo from Lama3 B2 is driven by an initiator element and an upstream TATA box.

The prototype pol II initiator was discovered in the promoter of the terminal deoxynucleotidyltransferase (TdT) gene [10]. When this initiator undergoes a double substitution to GA of nucleotides -2 and -3 (relative to the start site at +1), transcription is reduced by 80% (ref. 10). An identical double substitution in the DB fragment reduced luciferase synthesis in vivo by 75% (Fig. 2b, DB/GA). This indicates that the B2 initiator is functioning in a similar manner to prototype pol II initiators. Further support for this comes from the observation that transcription is maintained when the B2 initiator is substituted by the corresponding 10-bp sequence from the TdT promoter (Fig. 2b, DB/TdT). The resultant construct is half as active as the native DB construct, however, which indicates that the B2 initiator is very efficient in this context. Similar effects are seen when 90 bp of downstream sequences are removed from the DB fragment (Fig. 2b, DBstY/TdT).

Figure 3: B2 initiator sequences can be bound and activated by USF.

Figure 3: B2 initiator sequences can be bound and activated by USF.

a, EMSA using radiolabeled Lama3 B2 inr oligonucleotide probe and purified recombinant USF (lanes 2–7; ref. 23). Lanes 3–5 also contained 20-fold molar excess of unlabeled oligonucleotide containing the B2 inr sequence (lane 3), 50-fold molar excess of either E-box (CACGTG) sequence (lane 4) or a mutant B2 inr where (-3) TCAGAC (+3) was replaced by GGGAAA (lane 5). Lane 6 also contained antibody 63 against USF-1 (ref. 23) and lane 7 contained the unrelated YY1 antiserum.

b, The pol II promoter of the Lama3 B2 SINE is activated by USF in vivo. NIH-3T3 cells were transfected with the USF expression vector pRK-USF (ref. 24) or empty vector, along with CMV-_bgal and a luciferase reporter cloned downstream of the DBstY fragment of the Lama3 B2 SINE, B2 SINE fragments from Col18a1 or Tcrg, or the tyrosinase (Tyr) or SV40 early promoters. The fold stimulation shows the increase in luciferase production observed in the presence of pRK-USF.

c, USF1 activates pol II transcription from Lama3 B2 in its endogenous chromosomal location. NIH-3T3 cells were transfected with increasing amounts of pRK-USF1 (lanes 2–4) or with empty pRK as negative control (lane 1). Bottom, fold stimulation relative to Gapd, as an internal control.

d, B2 SINEs from the Col18a1 and Tcrg have initiator homologies that support pol II transcription in vivo. Top, NIH-3T3 cells were transfected with CMV-_bgal and a luciferase reporter subcloned downstream of the Lama3 DBstY fragment, or B2 SINE fragments from Col18a1 or Tcrg containing the sequences depicted in the lower part of this figure and extending further upstream. Values are given relative to the DBstY construct, which is assigned a value of 100. Bottom, sequence alignment of the initiator and TATA regions (underlined) of B2 SINEs from Lama3, Col18a1 and Tcrg, and the B2 consensus sequence.

The helix-loop-helix factor USF binds to initiator elements and stimulates expression from core pol II promoters [11, 12]. Electrophoretic mobility shift assays (EMSA) revealed that USF also recognizes the initiator of Lama3 B2. Thus, a complex was formed between purified recombinant USF and an oligonucleotide containing the B2 initiator (Fig. 3a). This complex results from sequence-specific binding, as it can be competed by a consensus USF recognition site, but not by a mutant initiator. Complex formation can be supershifted by an antiserum against USF (Fig. 3a), whereas it is not affected by an unrelated YY1 antiserum. To investigate the functional significance of this interaction, we co-transfected a USF expression vector into NIH-3T3 cells along with a luciferase reporter downstream of the DBstY fragment, which contains 70 bp of Lama3 B2 sequence including the TATA box and initiator elements. Co-expression of USF was found to activate this construct by 38- to 50-fold in vivo (Fig. 3b). This effect was highly specific, as USF produced little or no activation when co-transfected with reporters driven by the tyrosinase (Tyr) or SV40 early promoters. Endogenous Lama3 in its natural chromosomal context is also activated strongly by USF, in a dose-dependent manner (Fig. 3c). We conclude that the Lama3 B2 SINE contains a functional pol II promoter that is recognized by USF and can be strongly stimulated by it in vivo.

The 70-bp minimal pol II promoter delineated within Lama3 B2 has substantial nucleotide similarity with many other B2 SINEs (Fig. 3d). To determine whether Lama3 B2 is exceptional in its ability to direct pol II transcription, we inserted equivalent regions from two additional B2 SINEs upstream of the luciferase coding region; both were found to drive luciferase expression when transfected into NIH-3T3 cells (Fig. 3d). One of these SINEs was isolated from the T-cell receptor gene [13] (Tcrg) and has a poor TATA box (TTATGTTAT); its high activity is therefore consistent with our mutational analysis of Lama3 B2 (Fig. 2b), which showed that the initiator is the dominant contributor to function. The other SINE tested was derived from the collagen XVIII gene [14] (Col18a1); it matches closely the B2 consensus sequence and may be considered representative of the family as a whole. Not only can the Tcrg and Col18a1 promoters drive luciferase transcription in vivo, but they are also stimulated by 34- to 43-fold using the USF expression vector (Fig. 3b). The ability to direct USF-responsive pol II transcription is therefore common to several B2 SINEs. Database analyses revealed substantial homology between the three minimal B2 promoters tested and hundreds of published B2 sequences. It therefore seems likely that the properties we identified will be widespread within the B2 family.

Figure 4: The Lama3 B2 SINE contains a functional internal pol III promoter.

Figure 4: The Lama3 B2 SINE contains a functional internal pol III promoter.

a, Sequence alignment of the Lama3 B2 SINE with the consensus for the B2 family and the consensus sequences for the A- and B-block internal pol III promoter elements typical of tRNA genes. Also shown is the potential of the Lama3 B2 SINE to adopt a cloverleaf secondary structure typical of tRNA molecules; this is consistent with the evolutionary origin of the B2 family [2, 4].

b, Transcription in vitro using HeLa whole cell extract and AE template (lanes 1 and 2) or AemA, in which the
Lama3 B2 A-block sequence TGGCTCAGCGG has been replaced by CTCGAGAGATC (lane 3). Tagetitoxin was included in lane 2.

c, The Lama3 B2 SINE has a pol III promoter that is active in vivo. Normal human keratinocytes (NHK) were transfected with AE (lane 1) or AEmA (lane 2) along with CMV-b-gal as an internal control. Total RNA was
extracted and analyzed by primer extension.

Figure 5: Sequence alignment of the initiator and TATA box regions of the B2 (I) consensus [4, 5] with several Mus musculus pol II promoters from the Eukaryotic Promoter Database, for which the transcription start site has been determined experimentally (http://www.epd.isb-sib.ch/ ; SIECR, Lausanne).

Comparison was done with the LALIGN program ( http://www.ch.embnet.org/software/LALIGN_form.html ). Only significant stretches of at least 60-nt homologies are retained with opening gap penalties of -14 and extending gap penalties of -4. The transcription initiation sites are designated +1 and the first nucleotide is in bold. Homology scores are indicated in brackets.

Our evidence indicates that B2 SINEs have the potential to distribute a functional pol II promoter throughout the genome. During the course of evolution, this is likely to have resulted in many novel mRNA molecules, which would provide raw material for natural selection. A clear example of this phenomenon is provided by the Lama3 transcript [7]. Further examples may be provided by ESTs in the database, which begin at the pol II start site we have mapped in B2 genes (data not shown). In addition, several well-characterized pol II promoters show sequence homology with B2 genes over their core promoter regions (Fig. 5) and are therefore likely to be relics of a SINE insertion event. Although retrotransposition is often thought to be random, chromosomal hybridization has shown that SINEs tend to cluster around the R bands, where active genes are concentrated [16, 17]. This feature would have maximized the impact of their insertion. It is known that SINEs can influence expression of nearby genes [18-20], but we have presented the first evidence that these repeat elements can supply a pol II promoter. We propose that the B2 family has contributed in a novel way to evolution, by serving to distribute functional and regulatable pol II promoters.

Methods:

Plasmids and site-directed mutagenesis.
Plasmid AE has been described [7]. Constructs AB, DE, PB and DB were generated by digesting AE with, respectively, AflII/BlpI, DraI/EagI, PvuII/BlpI and DraI/BlpI. Each resulting gel-extracted insert was blunted and subcloned into SmaI-digested pGL2-basic vector. DB plasmid was digested with BlpI/BstYI, blunted and self-ligated to obtain DbstY construct. To obtain AEDDB, AE construct was digested with DraI/BlpI, blunted and self-ligated. Col18a1 and Tcrg B2 fragments were PCR-amplified with, respectively, the primers
                  5'–TCAGTATCAGTTGGCTCTG–3' and
                  5'–GCCCACTAGTTCCCTCTCC–3', and
                  5'–AGCCACTGAAGCATCTCTC–3' and
                  5'–GAGCTGGTCCCTGGGGCAC–3'.
The resulting 375- and 370-bp fragments were ligated into the TOPOII cloning vector (Invitrogen), digested with KpnI/BstYI, and the 145-bp and 243-bp inserts were ligated with KpnI/BglII-digested pGL2-basic vector. Sequence changes were introduced into the TATA box and initiator regions of the DB fragment of the Lama3 B2 SINE using the Quickchange site-directed mutagenesis kit (Stratagene). Substitutions were confirmed by DNA sequencing.

Cell lines and transfections.
We cultured and transfected NIH-3T3 fibroblasts and normal human keratinocytes (NHK) as described [7, 21]. CMV-_b-gal was co-transfected to control for transfection efficiency. Values shown are the means of triplicates from three independent experiments and have been normalized for _b-galactosidase expression.

Primer extension analysis.
We carried out primer extension using total RNA prepared from NHK with primer
                  5'–ACAGTTTTGTGAGCCACCGTGTGGTTG–3' for B2 or
                  primer 5'–AGCAGGCTCTTTCGATCCCCAAGC–3' for
                 _b-galactosidase as described [7].

In vitro transcription assays.
In vitro transcription assays were conducted as described [22] using HeLa whole cell extract (30 _mg) or a reconstituted system containing purified pol II, TFIIA and TFIIH along with recombinant TBP, TFIIB and TFIIE. When indicated, a-amanitin (0.2 _mg/ml) or tagetitoxin (10 _mM; Epicentre Technology) were included.

EMSAs.
Mobility shift assays with purified recombinant USF (PET-USF; ref. 23) were performed as follows: the protein was incubated for 20 min at 30 °C with 0.1–0.2 ng of radiolabeled Lama3 B2 inr oligonucleotide (5'–CATTGCTCTCTTCAGACACACCAGAAG–3') as probe (in a buffer containing 20 mM Tris-HCl, pH 7.9, 10% glycerol, 10 mM EDTA, 5 mM dithiothreitol (DTT), 80 mM KCl, 200 ng of poly(dI.dC) as nonspecific DNA and bovine serum albumin) to bring the total protein concentration to 500 ng, in the presence or absence of USF or YY1 antibody (Santa Cruz) and with or without specific competitors, as indicated. The reaction products were analyzed by electrophoresis in 6% non-denaturing polyacrylamide gels. The radioactive signals were visualized by autoradiography.

RT–PCR.
Poly(A)⁺ RNA was isolated directly from NIH-3T3 cell line using the Micro-Fast-Track 2.0 mRNA Isolation kit
(Invitrogen). Poly(A⁺) RNA from mouse lung was obtained from Clontech. DNase-I-digested poly(A⁺) RNA was reverse transcribed using primer 5'–CGTAATTCTCTGCAGGTACCACTAGAACC–3' for the Lama3 transcript. The products of the reverse transcription reactions were denatured for 5 min at 94 °C and one-tenth was subjected to amplification by 30 cycles of PCR at 94 °C (45 s), 60° C (45 s) and 72° C (45 s) for elongation. The PCR products were then separated on 2% agarose gels. Control PCR amplification in which the reverse transcriptase was omitted has been systematically done for each couple of primers. Location and sequences of the primers used are indicated in Fig. 1c. Each amplified product was cloned and sequenced.

NIH-3T3 cells were transfected with increasing amount of pRK-USF1. Total RNA (1 _mg) recovered from transfected cells was reverse-transcribed using primer
                  5'–TTGTGAGGCCAGACACT–3' for the pol II Lama3 B2 transcript. Equal amounts of cDNAs were subjected to 25 cycles of PCR reaction with specific primers
                  (5'–GTCCTCTACACTGCATGTTA–3'/
                  5'–GGGAAGCAGCACCAGGTAGTCCAGAAGGAC–3') and the products were resolved
                  by 1.5% agarose gel electrophoresis. The gel was Southern blotted and hybridized with the primer
                  5'–CTGGACTGCCTCACAG
                  ACAATCTCACCCTTACCTT–3' as specific probe for the Lama3 transcript. Gapd was amplified in parallel as internal control for RNA samples.

We thank J.M. Egly and F. Coin for pol II transcription factors and HeLa whole cell extract; P. Pognonec for pRK-USF1 and pET-USF; G. Pages for pVEGF; and P. Rigby for critical reading of the manuscript. This work was supported by grants from INSERM, Association pour la Recherche sur le Cancer (contract 9657) and EEC BIOMED 2 (BMH4-97-2062).

SINE retrotransposons such as B2 and Alu have an evolutionary history of insertion into new chromosomal loci, resulting in the formation of new DNA and new promoters, ie. new genes. Such new genes have been stably conserved for over 5 to 6 million years, since before the separation of human and chimpanzee evolution. Are these mobile sequences still active in the formation of new genes ?

1. Otieno AC, Carter AB, Hedges DJ, Walker JA, Ray DA, Garber RK, Anders BA, Stoilova N, Laborde ME, Fowlkes JD, Huang CH, Perodeau B, and Batzer MA, "Analysis of the Human Alu Ya-lineage", J. Mol. Biol., vol. 342, no. 1: pp. 109-18, (September 3, 2004).

2. Salem A-H, Ray DA, Xing J, Callinan PA, Myers JS, Dale J, Hedges DJ, Garber RK, Witherspoon DJ, Jorde LB, and Batzer MA, "Alu elements and hominid phylogenetics", Proc. Natl. Acad. Sci. U.S.A., vol. 100,  no. 22, pp. 12787–12791, (October 28, 2003). ( DOI: 10.1073/pnas.2133766100 ).

3. Allen TA, Von Kaenel S, Goodrich JA, and Kugel JF, "The SINE-encoded mouse B2 RNA represses mRNA transcription in response to heat shock", Nature Structural Molecular Biology, Published online: 08 August 2004; | doi:10.1038/nsmb813

4. Espinoza CA, Allen TA, Hieb AR, Kugel JF, and Goodrich JA, "B2 RNA binds directly to RNA polymerase II to repress transcript synthesis",  Nature Structural Molecular Biology, Published online: 08 August 2004; | doi:10.1038/nsmb812

Links to Euchromatin Activator RNA Reviews:
Links to Euchromatin Activator RNA Research:
Links to Ultrastructural Probes of DNase I-Sensitive Sites:
Links to RNA as a Therapeutic Agent:
Links to Hodgkin Lymphoma Immuno-Pathology:
Links to Activated T-Lymphocyte Immunotherapy:
Links to Medical Systems Biology:
Links to Selective Gene Transcription:
Links to RNA-Induced Epigenetics:
Links to RNA-Induced Embryogenesis:
Links to RNA and Biological Causality:
Links to Reprogramming and Neoplasia:

"Ultrastructural Probes of Active DNA Sites, and the RNA Activators of DNA".

For Further Information and Feedback:
E-mail: frenster@euchromatin.net