Terminal DNA

Archetypal Telomeres

The telomeric sequences of many linear chromosomes and plasmids of Streptomyces are available. The telomeric sequences of most Streptomyces linear plasmids characterized so far resemble those of Streptomyces chromosomes. The term "archetypal" is used here to describe this group of conserved telomeric sequences, which need to be distinguished from a number of atypical telomeres (such as SCP1, pRL1, pRL2, and the S. griseus chromosome; see

A. Archetypal telomeres

SAP1 SAP2 pSV2(r) pFRL1 PSCL1 pSLA2 S. lipmanii S. rimosus

S. lividans

S. coelicolor S. scabies S. avermitilis(/) S. avermitilis(r) pSCL2 ». cinnamoneum

S. parvulus pSPA1 SLP2(/) pSLA2-L

20 Tap binding 40_Tap binding 6,0__80 1(J0

EEEEEEEEEEEEE AE^AEATE^iE^^ATET^QAE' feGGGCGG-TOEEHEEEETECCGCTG -CC^ACGSGaiGCET^ECGGATCCCiiiCECTGE^

t EHJEEEEtEc gctg -c EacgS ¡Eg ! t eecggat ccQ cEctgEEE

^ AggAgATg^ilcgcgcATgT^gAg gG GCGG-T grSircgrcgTgC GCTGGCCgTCGS giG )\ A g TGGTC CTll CgCTGT

S^ETAEEABATEaaiEBsBsATETaaBAE^aagcggctcEBj|S[StEc cg cg -cc5gtgH Wg c g cggtc ccn cgctgi

_S^ETAEETQGGEmiE^^CC||A^EAE|ECATGCCCC EH1EISLS&3GGGCA- -G SACGE S|G A E ---AGCCCB AEGCGt.__

^{^ETcEETEcGEESHlt3E!33cGEAE3SGG[^CACGcccc EHJEEESGSGGGCGT-G HACGE -HEGI A E CCCGC CTE GEGCGEEE

HsHEHEEE HsIiEIiEEE HsHEHEEE

HsHEHEEE

u rggjllcgcra rEBSEEBSB

g[^gTAggAgATg^IIgggggATgTgggAg[jgAACACCCG ggi^^AgG g^g AggAgATg^^^^ATgT^QAg gAACACCCG gffl^^AgG

EESETAEEaEatESIE®^ATETSQAEiSAACAccCG EHJEEEISGEG tTEISiii^^ATET^0AE^AACCACCGCEHJE~ ~

VCg^^^^GTgG^gAg gAACACCCG gfflg

CEAG-GATECGGTGGGGCCGGCC^S 03iGGC-GCGG 136 CEGG- GATECGGTGGGGCCGGCCEHi EHfGAC- - CGG 135 ^gAG-GGAgCGGTGAGGCCGGCCTcTSl gj3jACC- - CGG 136 WTG - GGAWCGGTGGGGCCGGCCwS»G[»ln»GCC - - CGG 136 lEGTTGGGgCCGCTGCGCGGGCC^g-S|EtGC-GCTG 134 SGT T CGGEC - GCTGCGCGGCCTEEE CC -G CCC 133

ffGCGTCCgGC-TCACCAACCCGfcTSj g^CG - -GGTT 135 gGCGTCCWGT-TCACCAACCCG^fflagWCG - -GGCT 135 aGCGTCCffGC-TCACCAACCCGpTili peRCG - -GGTT 136

¡£ATgGgjg-g gA CACCCGDggi^^GgGGGTGTTG igCTCS ^T G g GAGCGCTG ?ATgG^g-g gA CACCCGffiffl^^GgGGGTGTTGcScTcfflc^T G g GAGCGCTG^GP-GAPI

___________¿ATQG^QAE Ea CACCCG "CTMSMSGWGGGTGTTG ^CTCBCBST G E GAGCGCTGE^GE-GAEI___ _ ^^__

sEEEE aEEaEatEEEHESESEatEtEEEaE Ea cacccg EHiEEEEcEGGGTGTTG IctcE EEt g E gagcgctgE gE-gaEEE ^gcgtccEgc-tcaccaagccc^StE3sgg ■ -gctt 136

^GGTGTTGCgCTTgC^TCgG^gAAGCGCTGggGgTGGQ^il^GCCCATgGCAGCAACTCGATTr^iag^AATCGAGT 135 gGGGTGTTGCaTTCgC^TCc G P GAACGCTG®GgGGAH^nGgGCATCCWCGCTCACCACCATT^fflGgWAAT --GGT 138

_________________________SGGGTG TTG SlCTcScgffTC G F GAGCG CGF GE - GAffEwnGEiGCGTCCffGC - TCACCTCGGTTEflflaPltpAACCGGG - 137

«EEEE AEEAEATEEEiliHEHEATETEEnAE EA CCACCGCEHJSJ^GEGGTGGTTG actce EET g E gagcgctgf ge - CCEET»n(^gcgggcpac - TCACCCCCGGTPT31aEra»accaggg ■ 137

gAACACCCGTgrargggAgGGGTGTTGCgCTCgCggTC G g GAGCG AAg Gg - GCggincgGCGGCCgGC - - - ACCCCCGGTgMlaggsATCGGGG - 135 gA CACCCGCgffl^^GgGGGTGTTG CgCTCfflo^TC c G ggnnGAGCGTGAggTg - CTg^O(^GCGAGCOGC - TCACCCGCCGOEWITP^AC- -GGCG 136

___________SATEG^QAE £A CACCCG E^EEESGEGGGTGTTG 3cTcEc0STCyGa£iGAGCGOCcEgTE -CCE^UGEGCGAGCEGC - CCACCGGCCCG^ffleEasCG - -GGCC 136

sEEEE CEEcEatEEEHEHEHEAtEgEEQ-E EA CGAATT ESIEEEEgEATTCAA-- GAACE SET G E - - GTTTTT^AE-GGE^QTEGCGCCTDTCTTGAATTCACGC^EaESJGAATTCGG 133 iE^E gA CCGGGGC^ai^^GECCCGGT- - cICTCff [cT»T ; G F GAGCGGAGF1 Gtc - CT[»[tT»i1cr»GCTCACf»GCCAGGCGGGACGCfcT>fl PcT»G CGCCCCG 136 j i_i i_ii_i i_i i_

Telomeres Coelicolor

below). The archetypal telomeric sequences are most conserved over approximately the first 40 nt, and diverge increasingly away from this point (Fig. 1a). In most archetypal telomeres, the first 13 nucleotides form a palindromic sequence (Palindrome I). Palindrome I is followed by several tightly packed palindromes, some lying next to each other without interruption. In many (but not all) cases, another (usually the fourth) palindrome is complementary to Palindrome I in sequence.

When the 3' strand of the telomere sequences is folded into a secondary structure using an energy optimization algorithm, the most striking feature is the "rabbit ear" (or "Y-shaped") structure formed by the first four palindromes (with paired Palindromes I and IV; Fig. 1b). The general shape and loop sequences of this part resemble those at the 3' end of autonomous parvoviruses, which are single-stranded DNA molecules flanked by stable duplex hairpins at both ends (Astell et al. 1979). Autonomous parvovirus-like loops are also found in other hairpins of the folded archetypal Streptomyces telomere.

Most of the hairpins are closed by the YGNAR sequence (loop underlined; N being most often C). In such hairpins, the G and A residues have been shown by 13C NMR studies to form "sheared purine-purine pairing", resulting in a single C-residue loop (Zhu et al. 1995; Chou et al. 1997). Interestingly, GCA hairpins are the most stable among all GNA hairpins with a sheared G: A pair. These single-nt hairpin loops closed by sheared purine pairs resist single-strand specific nuclease attacks (Chou et al. 1997). The overwhelming presence of GCA loops at the 3' overhangs strongly implies their importance in shaping the 3' overhangs for structural integrity during replication, and possibly for the proper conformation for end patching.

A Fig. 1 Streptomyces telomere DNA. a Aligned archetypal telomere sequences. For the telomeres of plasmids, the plasmid names are given, and for those of the chromosomes, the species names are given. When the left ("l") and right ("r") terminal sequences of a replicon are different, they are distinguished in parentheses. The degree of relative conservation is indicated by the shading of the letters and the background. The lightest letters with the darkest background indicate the most conserved sites, and the darkest letters with the lightest background indicate the least conserved sites. Six palindromes (I-VI) are indicated based on the S. lividans chromosome (and SLP2 right end) sequence. These palindromes are preserved in most but not all the archetypal telomeres. The binding regions based on the pSLA2 sequence of Tap (Bao and Cohen 2003) are marked. b Predicted secondary structures formed by the 3' strands of the S. lividans chromosome. The Roman numerals indicate the palindrome numbers. The proposed Pu: Pu sheared pairings are indicated by the solid dots. The hairpin structures identical to those in the autonomous parvoviruses are shaded. c Secondary structures formed by the 3' strands of a pseudotelomere (nt 3974-4234) found on SCP1. The long stem structure is omitted. d Secondary structures formed by the 3' strands of the SCP1 telomere. Note the unpaired 3' end and the 4-nt loops

Nonarchetypal Telomeres

Four nonarchetypal Streptomyces telomere sets have been characterized so far: three are on the plasmids, and one on a chromosome. The most striking set is that of SCP1. It differs remarkably from the archetypal telomeric sequences in primary structures and potential secondary structures formed by the 3' overhangs (Kinashi et al. 1991; Bentley et al. 2004). It begins with a string of four to six Gs instead of three Cs in the archetypal telomeres. In the predicted secondary structure of the SCP1 telomere sequence, the 3' end is unpaired, in contrast to the fully duplexed 3' ends formed by many archetypal telomeres (Fig. 1d). Moreover, all the loops consist of four rather than three nucleotides at the archetypal telomeres. Most of the 4-nt loops may still form purine-purine sheared pairing between the first and fourth nucleotides. These distinct features hint at a very different capping TP for this plasmid (see below).

The telomere sequences on two linear plasmids, pRL1 and pRL2 (Zhang et al. 2006) and on the S. griseus chromosome (Goshi et al. 2002) also share no homology in primary and secondary structures with either the archetypal telomeres or the SCP1 telomere. Although multiple palindromes are present in these telomere sets, they are relatively sparse and simple compared to those in the archetypal telomeres. Unlike the SCP1 telomere, the predicted secondary structures of these telomeres contained 3-nt instead of 4-nt loops.

Terminal Inverted Repeats

The telomeres on a linear plasmid (or a chromosome) of Streptomyces are often part of longer terminal inverted repeats (TIRs), i.e., identical terminal sequences. The sizes of most TIRs range widely from tens of base pairs (bp) to hundreds of kilobases (kb). The sequences of the TIRs are also very heterogeneous with few constraints, except for the telomere sequences and perhaps a terminally located helicase gene ttrA that is highly conserved (Bey et al. 2000).

The size and sequence variations turn out to be the results of intermolecular exchanges that are frequent enough to be observed in the laboratory (reviewed by Chen et al. 2002). In general, TIR shortening may result from internal deletions or from recombination between heterologous linear replicons (with different terminal sequences). This is illustrated by the chromosomes of S. coelicolor-S. lividans hybrid strains produced by conjugation, which contain a telomere from each parent and exhibit imperfect TIRs of less than 200 bp (Wang et al. 1999). The 44-bp imperfect TIR of the SLP2 plasmid (Chen et al. 1993) and the 174-bp TIR of the S. avermitilis chromosome (Ikeda et al. 2003) were presumably generated in a similar fashion. More recently, Ya-

masaki and Kinashi (2004) discovered two chimeric chromosomes produced by recombination between SCP1 and the host chromosomes in S. coelicolor 2106, each of which contains a telomere of SCP1 and a telomere of the chromosome. An artificially constructed linear plasmid containing this mixed pair of telomeres can also replicate in vivo (Huang et al. 2007). These two telomeres share no homology down to the terminal nucleotide (C for S. livi-dans chromosome and G for SCP1), and therefore the recombinant plasmid and the 2106 chimeric chromosomes have no TIR. Thus, TIRs appear to have no significant biochemical function, although the duplication may serve as a template for repair of damage in the TIR (Qin and Cohen 2002), and so provide an evolutionary advantage.

On the other hand, TIR may also be increased by unequal crossovers between sister replicons, as was observed by Fischer et al. (1998) in S. am-bofaciens and Uchida et al. (2003) in S. griseus. The unusually large (1-Mb) TIR found in some strains of S. coelicolor (Weaver et al. 2004) may also be the result of such an event.

Pseudotelomeres

"Pseudotelomeres" (Fig. 1c) were originally discovered as internal sequences of SLP2, which exhibit remarkable similarity to archetypal telomeres in both primary and secondary structures, such as the Y-shaped "rabbit ear" structures and the stems with 3-nt (often GCA) loops (Huang et al. 2003). Subsequently, two pairs of pseudotelomeres were also identified in the TIRs of SCP1 (Bentley et al. 2004). These pseudotelomeres lie near the end of both plasmid DNAs (at 1.6 kb on SLP2 and at 4.0 kb on SCP1). It is interesting that these pseudotelomeres all occur in pairs (and in the same sequence orientation) and are separated from each other by only 1.8-2.1 kb. They are presumably the remnants of integrated linear replicons. It is not known whether these "pseudotelomeres" are still functional in Streptomyces when placed at the end of linear replicons.

Recently, on a subclone of SLP2, Xu et al. (2006) discovered a 287-bp internal deletion corresponding to 91% of one of the predicted pseudotelom-eres. Perhaps these pseudotelomeres readily form secondary structures under negative superhelicity in vivo, which are prone to either nucleolytic attacks or replication slippage. A more speculative alternative would be that these elements are mobile.

0 0

Post a comment