Maize Genetics Cooperation Newsletter 80. 2006.

Pascani, Republic of Moldova

Maize and Sorghum Research Institute

Bg transposon: a possibility of regulation of transcription through formation of Z-DNA and Z-DNA binding properties of its encoded proteins

--Koterniak, VV

Probable structure and properties of Bg-encoded proteins. Previously, on the basis of Bg transposon sequence, the primary structure of its putative encoded proteins (designated as PPBg1, PPBg2 and PPBg3; MNL 79) was proposed. The sequence of one of them, PPBg3, consists in fact of sequences of the two other proteins (PPBg1 at the N- end and PPBg2 at the C-end) divided by an insertion of 38 amino acid residues (MNL 79). Further analysis of this protein's sequence indicates that besides domains described earlier (DNA binding, catalytic activity, oligomerization; see MNL 79) it contains 2 regions similar to nuclear transport signals. The first one includes residues R13-R15 and H28-K32 (highlighted in bold in the sequence RRRSNATVTDEQDDCHRKGK; see Fig. 1c in MNL 79) and resembles a nuclear localization signal (NLS) motif of the basic bipartite type (Yoneda et al., Cell Struct. Funct. 24:425-433, 1999; Xiao et al., J. Biol. Chem. 276:39404–39410, 2001). The structure of the second region (LxxxLxxLxL, residues L396-L405 of the sequence LVVALQFLVL; see Fig. 1c in MNL 79) is similar to the nuclear export signal (NES) of MAP (mitogen-activated protein kinase kinase) (Yoneda et al., 1999; Xiao, 2001; Perander et al., J. Biol. Chem. 276: 13015–13024, 2001).

Indicated NLS and NES signals are encoded by the two largest ORFs of the Bg transposon, located near its 5' and 3' ends. (For PPBg3 these ORFs constitute the first and the last exons and encode 84% of its sequence.) It is necessary to mention that nuclear transport signals are characteristic for large (e.g., PPBg3-like, the molecular mass of which is 69.4 kDa) proteins (see for example Yoneda et al., 1999). All this, even in the absence of experimental data on Bg-encoded proteins, strongly suggests that at least one of the products of this transposon should be a large protein, and that the two longest ORFs of Bg take part in encoding this protein sequence.

Similarity of a region in the N-end of Bg-encoded proteins to the Za domain of Z-DNA binding proteins. Quite unexpected was the resemblance of a region in the N-end of probable Bg-encoded products PPBg3 and PPBg1 to the Zα domain of DNA binding proteins (Fig. 1). This region is encoded by a part of the longest 5' end ORF starting from the 813 position of Bg sequence (numbering of Bg element bases is given according to its GenBank accession number, X56877.1). In Figure 1 the comparison between PPBg3 and double-stranded RNA adenosine deaminase (ADAR1, showing the highest affinity of Z-DNA in comparison with other Z-DNA binding proteins used; see Kim et al., PNAS 101:1514-1518, 2004) is highlighted. It is necessary to mention that the part of the Zα domain used contains 10 out of 17 amino acids which are important for the protein fold and for Z-DNA recognition (see Kim et al., PNAS 100:6974–6979, 2003).

PPBg3 71 VNKKSLFMVLYSCIFKILWSYTAG 94

hZα_ADAR1167 TPKKEINRVLYSLAKKGKLQKEAG 190

mZα_DLM140 VPKKTLNQVLYRLKKEDRVSSPEP 62

Yaba 41 INKKKINQQLYKLQKEDTVKMVPS 64

Vaccinia 38 MEKREVNKALYDLQRSAMVYSSDD 61

Figure 1. Similarity of the region V71-G94 of PPBg3 to the Zα family of Z-DNA-binding domains. Sequences used according to GenBank accession numbers (designated on the Figure 1 according to Kim et al., 2003) are as follows: AAB06697.1, the double-stranded RNA adenosine deaminase (ADAR1, hZα_ADAR1, (Homo sapiens)); NP_067369, Z-DNA-binding protein 1, tumor stroma and activated macrophage protein DLM-1 (mZα_DLM1, Mus musculus); NP_073419, the 34L protein (Yaba, Yaba-like disease virus); AAA02759, the E3L protein, (Vaccinia, vaccinia virus). Comparison between PPBg3 and ADAR1 is highlighted. Identical residues are shown on a black background, similar ones are on a gray background.

In connection with the finding of the Z-DNA binding motif in Bg-encoded products and the known ability of transposons for autoregulation (see for example Raizada et al., Mol. Genet. Genomics 265:82–94, 2001; Kunze and Weil, Pp. 565-610 in "Mobile DNA II", Craig et al. (eds.), ASM Press, Washington, 2002.) it is tempting to find out whether regions of the Bg transposon's sequence are able to form Z-DNA.

Several regions in the 5' end of the Bg transposon sequence may potentially form Z-DNA. Analysis for the presence of Z-DNA forming regions in the Bg sequence was carried out using the ZHunt program (Ho et al., EMBO J. 10:2737-44, 1986; Ho, PNAS 91:9549-9553, 1994; Champ et al., Nucl. Acids Res. 32:6501-6510, 2004). (Access to this program was kindly offered by Prof. P. S. Ho from Oregon State University.) This analysis revealed two regions with Z-DNA forming potential: 1) positions 120-140 (accagacgcgcgcacgagagc, Z-score 2.2∙10⁴); 2) positions 402-414 (cacggacgcgcag, Z-score 9.3∙10²) (Fig. 2, see also Table 1). For convenience, they will be further referred to as BgZDR120 and BgZDR402, respectively (for Z-DNA forming regions, ZDR; Champ et al., 2004).

These regions are present upstream of the translation initiation start site for PPBg3 (and PPBg1, position 813 of the Bg sequence) in the 5' end of the Bg element characterized by high G and C content (lacking the TATA promoter). This G/C rich region resembles the similar one of the Ac transposon (Kunze et al., EMBO J. 6:1555-1563, 1987) and some mammal housekeeping genes (see Hartings et al., Mol. Gen. Genet. 227:91-96, 1991; Maydica 36:355-359, 1991).

The possibility of the presence of other regions with Z-DNA forming potential in the Bg sequence, besides the ones found by the ZHunt program, cannot be excluded. Thus, taking into account the pattern of distribution of purine and pyrimidine bases in Z-DNA forming sequences (see for example McLean et al., PNAS 83:5884-5888, 1986; Schroth et al., J. Biol. Chem. 267:11846-

11855, 1992; Herbert and Rich, J. Biol. Chem. 271:11595-11598, 1996; Champ et al., 2004) several regions in the Bg sequence (e.g., sequences found near BgZDR402 region, positions 388-423, cgcgggccggcgccacggacgcgcagcgccagggc; downstream of this region in positions 579-592, tggggtgtgccgcg and positions 674-686, gtgtggctgctgt) could be distinguished as having the propensities for forming Z-DNA. It is possible that these regions may form Z-DNA in the presence of certain specific cellular factors.

It is known that Z-DNA can be generated by transcription and that distribution of Z-DNA is nonrandom, with its preference for localization near transcription initiation sites (Wittig et al., PNAS 88:2259-2263, 1991; Schroth et al., 1992; Champ et al., 2004).

Transcription start site for Bg transcripts. Works on transcription initiation sites for Bg-encoded products are unknown to the author. However, taking into consideration the conserved motifs of initiator (Inr) and downstream promoter elements (DPE), the sequences which commonly are present in TATA-less promoters (see for example Zhang and Dietrich, Nucl. Acids Res. 33:2838-2851, 2005; Kadonaga, Exp. Mol. Med. 34:259-264, 2002; Nakamura et al., Plant J. 29:1-10, 2002), at least two motifs similar to the Inr elements located in good agreement with the DPE resembling motifs could be distinguished in the 5' Bg transposon end (Fig. 2).

One of these motifs is situated near the BgZDR120 region, the other one is near the BgZDR402 region, confirming the above-mentioned predisposition of Z-DNA for localization near transcription initiation sites. In the first Inr-like sequence (positions 116-122, cca₊₁aacc), the last three base pairs enter in the BgZDR120. The G-139 of the sequence similar to the DPE element (positions 139, 143-147, G₊₂₂, ggttt) is also a part of the BgZDR120 and is situated 2 base pairs (bp) closer to A-118 of the first Inr-like sequence in comparison to the consensus DPE motif of Drosophila (g₊₂₄, (g/a)₊₂₈g(a/t)(c/t)(g/a/c); Kadonaga, 2002). The second Inr-like sequence (positions 423-429, cta₊₁cttc) is located 9 bp downstream of the BgZDR402 region and is followed by the DPE-like motif of g₊₂₄gcgagtaC (positions 448-456).

Possible mechanisms of regulation of Bg transcription through the interaction between Z-DNA-binding domains of its encoded proteins and its Z-DNA forming regions. Formation of Z-DNA can activate transcription and act as the cis-element in genic regulation (Liu et al., Cell 106:309-318, 2001; Sheridan et al., Mol. Microbiol. 40:684-690, 2001; Oh et al., PNAS 99:16666-16671, 2002). Z-DNA binding proteins can stabilize Z-DNA and act as potent effectors of gene expression (Oh et al., PNAS 99: 16666-16671, 2002). On the other hand, the inhibitory action of Z-DNA on promoters is also known (Sheridan et al., 2001; Rothenburg et al., PNAS 98:8985-8990, 2001). Therefore, the determination of the real character of interactions between Z-DNA and Z-DNA-binding domains of Bg-encoded proteins and the effects of such interactions on the promoter activity need, of course, experimental studies.

However, taking into account that the activity of the Bg transposon can be assessed by the rate of excision of the nonautonomous rbg element from its mutable o2-m(r) alleles (leading to reversion of these alleles to the normal one), some assumptions could be made proceeding from features of the reversion of o2-m(r) alleles in the presence of Bg elements. Reversion of such alleles is characterized by the specificity of their interaction with different Bg transposons and can strongly depend on the dosage of these autonomous elements (Maydica 44:195-203, 1999; Maydica 48:275-281, 2003; Genetika (Moscow) 39:769-774, 2003). Two features of observed dosage effects can be underlined: 1) a significant, outstripping increase of reversion frequency of the o2-m(r) alleles when the dosage of Bg transposons increases from 1 to 3 or from 1 to 2; 2) an insignificant change in reversion frequency of the o2-m(r) alleles when the dosage of Bg transposons increases from 2 to 3 (see for example the behavior of the o2-hf allele in the presence of Bg-hf; Maydica 44:195-203, 1999; Maydica 48:275-281, 2003; Genetika (Moscow) 39:769-774, 2003).

Previously, several suppositions were made about the properties of Bg-encoded product(s) that explain the above-mentioned features of Bg dosage effects on the posttranscriptional level of its encoded products (Maydica 48:275-281, 2003; Genetika (Moscow) 39:769-774, 2003). However, finding the Z-DNA binding domains in Bg-encoded proteins and the ability of certain regions of Bg sequences to form Z-DNA may indicate the existence of mechanisms of regulation of Bg dosage effects on another, transcriptional level.

Thus, assuming that the formation of Z-DNA near transcription start sites enhances the efficiency of transcription by maintaining open confirmation of chromatin in this region (see Liu et al., Cell 106:309-318, 2001) and that there is a positive dependence between frequency of rbg excision and concentration of Bg-encoded proteins, a positive autoregulation mechanism for transcription of the genes encoding Bg products can be proposed. By this mechanism, transcription efficiency is enhanced by the increased stability of Z-DNA due to an enhancement in the stabilizing action on Z-DNA of the Z-DNA binding Bg-encoded proteins when the Bg dose increases from 1 to 2 or from 1 to 3.

Several explanations are possible concerning the role of Z-DNA located downstream of transcription start sites in the Bg sequence. The presence of these Z-DNA regions may increase fidelity of RNA splicing by a mechanism proposed by Wittig et al. (1992). Another explanation can be prompted by the aforementioned feature of an insignificant change in reversion frequency of the o2-m(r) alleles when Bg dosage increases from 2 to 3. For example, binding of Bg-encoded proteins at their enhancing concentration (when the dose of Bg elements increases from 2 to 3) to the indicated Z-DNA regions (especially if these regions show lower affinity to Bg-encoded proteins and are bound by such proteins at their high concentration) may hinder the movement of the next RNA polymerase molecule on these regions, thus lowering the efficiency of transcription. That is, the stabilizing action of Bg-encoded proteins on Z-DNA situated downstream of transcriptional start sites would affect gene transcription in a negative autoregulation mode.

In any case, the presence of Z-DNA binding domains in Bg-encoded products, and the ability of certain regions of the Bg transposon to form Z-DNA, may indicate the existence of mechanisms of Bg activity autoregulation through the interaction of Z-DNA forming regions of this transposon with its encoded proteins.

Z-DNA forming regions in other maize transposons sequences. Using the ZHunt program a search for Z-DNA forming regions in sequences of other maize transposons (Ac, PIF, En, MuDR) was carried out. Such regions were found in sequences of all transposons analyzed except PIF (GenBank accession number AF412282.1) (Table 1).

Table 1. Probable Z-DNA forming regions in sequences of different maize transposons revealed by the ZHunt program (Ho et al., 1986; Ho, 1994; Champ et al., 2004).

Transposon	Starting position	Length, bp	Z-score	Sequence
Bg	120	21	2.2×10⁴	ACCAGACGCGCGCACGAGAGC
	402	13	9.3×10²	CACGGACGCGCAG
Ac	381	16	3.3×10³	CCACGCGCCCACGCCG
	1261	28	1.4×10⁵	ATGTACGTGCACGTGCGCGTGGGCATGG
En	397	15	1.8×10³	GAGCGCGCACCTCCA
	5892	13	4.7×10³	TTCGCGTGTGCGA
	8035	17	2.4×10³	TGATGTGCGCGCAGTAA
MuDR	169	19	1.8×10⁴	TTCGCCCGCGCACACGCCG
	4756	20	1.8×10⁴	CGGCGTGTGCGCGGGCGAAC

Sequences used (according to the GenBank accession numbers) are as follows: X56877.1 (Bg); X05424.1 (Ac); M25427.1 (En); M76978.1 (MuDR). The Z-Score cutoff (minimum) is equal to 700.

Interesting results are observed for potential Z-DNA forming regions of the MuDR transposon. A characteristic feature of this transposon is the convergent transcription of its two major transcripts, mudrA and mudrB, initiated in terminal inverted repeats from opposite strands (Hershberger et al., Genetics 140:1087-1098, 1995). Transcription start sites of these transcripts (Hershberger et al., 1995) are located near revealed Z-DNA forming regions: the beginning of the first Z-DNA forming region (starting from position 169 of the MuDR element sequence, see table 1) coincides with the starting bp nucleotides of the first start site of the mudrA transcript; the second Z-DNA forming region (positions 4756-4775 of the MuDR element sequence, see Table 1) is located 5 bases downstream of the transcription start site of mudrB (position 4780; Hershberger et al., 1995). These results confirm one more time the predisposition of Z-DNA for transcription start sites and indicate the involvement of Z-DNA in the regulation of the transcription of the MuDR transposon genes.

Corrigendum. In the MNL 79 note on Bg-encoded proteins a misprint was made in the legend of Figure 1: the correct numbers for the first exon of PPBg3 are 813-1546.

_________________________________________________

Please note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors.

Return to MNL 80 on-line index.

Return to MNL index.

Return to MaizeGDB home page

____________________________________________