Fig.7.2 RNA polymerase II core promoter elements. DNA

sequences prevalent within the core promoters of RNA polymerase II transcribed genes have been identified. These sequence motifs include the TATA box, initiator (INR), downstream promoter element (DPE), and motif ten element (MTE), all of which serve as binding sites for subunits of the TFIID complex. The BRE, located immediately upstream of the TATA box in a subset of TATA- containing promoters, is a binding site for the general transcription factor TFIIB. Any given promoter will contain one or more of these core promoter motifs with the preferred combinations illustrated. The diversity of core promoter regions provides an additional level of complexity to transcriptional regulation in higher eukaryotes.

Fig.7.2 RNA polymerase II core promoter elements. DNA

sequences prevalent within the core promoters of RNA polymerase II transcribed genes have been identified. These sequence motifs include the TATA box, initiator (INR), downstream promoter element (DPE), and motif ten element (MTE), all of which serve as binding sites for subunits of the TFIID complex. The BRE, located immediately upstream of the TATA box in a subset of TATA- containing promoters, is a binding site for the general transcription factor TFIIB. Any given promoter will contain one or more of these core promoter motifs with the preferred combinations illustrated. The diversity of core promoter regions provides an additional level of complexity to transcriptional regulation in higher eukaryotes.

TAtA Box Element

The first eukaryotic core promoter element identified in protein encoding genes was the TATA box (Goldberg, 1979). The A-T rich sequence was discovered by comparing the 5' flanking regions of a number of eukaryotic and viral genes. In these early studies, the sequence TATAAA, now considered the canonical TATA sequence, was present in nearly all the RNA polymerase II transcribed genes that were examined. Other studies later revealed that a wide variety of A-T rich sequences could function as a TATA element (Singer et al., 1990). These results lead to a modified TATA consensus sequence, 5'-TATA(A/T)AA(G/A)-3'.

In metazoans, the TATA box is typically found around 25 to 30 nucleotides upstream of the transcription start site. However, the position of TATA elements in yeast promoters is more variable, ranging from nucleotide -40 to -100 relative to the start site of transcription (Struhl, 1989). Mutation or removal of the TATA box has been shown to reduce or abolish the transcriptional activity of eukaryotic promoters in transient transfection experiments and in in vitro transcription assays (Grosveld et al., 1982; Grosveld et al., 1981; Hu and Manley, 1981; Wasylyk et al., 1980). These findings suggested that the TATA box is essential for transcription and would be present in all RNA

polymerase II transcribed genes.

With the sequencing of eukaryotic genomes and the identification of more and more core promoters, it has become clear that the prevalence of the TATA box is much lower than originally predicted. Computational analysis of 1941 Drosophila genes indicated that, between nucleotides -45 and -15, a match to the TATA consensus motif was found in 28.3% of the genes examined (Ohler et al., 2002). Expanding the region of analysis to nucleotides -60 to -15 only increased the percentage of TATA containing promoters to 33.9%. A similar database analysis of 1031 potential human core promoters revealed that a TATA box was present in only 32% of the genes (Suzuki et al., 2001). Therefore computational algorithms that use the presence of a TATA sequence as one of their criteria for promoter prediction will be ineffective for the majority of eukaryotic protein encoding genes.

A: TATA Box Binding

The TATA box is a binding site for the TATA binding protein (TBP), a subunit of the transcription factor IID (TFIID) complex (Hahn et al., 1989; Hoey et al., 1990; Peterson et al., 1990). TFIID was biochemically defined by the Roeder lab as an activity present in human HeLa cell nuclear extracts that is essential for transcription from TATA-containing promoters (Matsui T, 1980). DNase I footprinting experiments demonstrated that TFIID directly contacted the TATA box within the adenovirus major late promoter (Parker and Topol, 1984; Sawadogo and Roeder, 1985). Unexpectedly, it was very difficult to purify and clone TFIID from human or Drosophila cells. The activity appeared to reside in a large protein complex (Nakajima et al., 1988). Cloning of TBP was facilitated by the discovery that in Saccharomyces cerevisiae TATA binding activity resided in a single polypeptide (Buratowski et al., 1988; Hahn et al., 1989). The yeast TBP sequence was subsequently used to isolate homologues from other eukaryotes (Hoey et al., 1990; Peterson et al., 1990). Antibodies against TBP revealed that in human and Drosophila cells TBP is associated with other proteins ranging in size from ~20 kD to 250 kD, subunits of the TFIID complex (Dynlacht et al., 1991; Tanese et al., 1991). A detailed discussion of TBP and the TBP associated factors (TAFs) is provided in a subsequent article.

The binding of TFIID to promoter DNA is required to nucleate the assembly of the basic transcription machinery. It is now widely accepted that the majority of RNA polymerase II promoters lack a binding site for TBP. Therefore the recruitment of TFIID to TATA-less promoters cannot be driven by TBP-TATA box interactions and must involve other DNA elements within the core promoter, allowing for greater diversity in the mechanisms regulating transcription initiation.

Intiator Element (Inr)

The initiator is a core promoter element that encompasses the start site of mRNA synthesis (Smale and Baltimore, 1989). A comparison of efficiently transcribed protein encoding genes revealed a conserved adenosine at the +1 position and a conserved cytosine at nucleotide -1, surrounded by pyrimidines (Corden et al., 1980). Removal of the putative initiator element (Inr) within a number of eukaryotic TATA-containing promoters reduced transcription levels and increased the heterogeneity of the position of transcription initiation (Concino et al., 1984). These findings suggested that sequences within the vicinity of the transcription start site can direct the location and level of transcription mediated by RNA polymerase II from the core promoter.

The importance of sequences around the transcription start site was further strengthened by a comprehensive analysis of the murine terminal deoxynucleotidyltransferase (TdT) promoter, a TATA-less promoter that initiates transcription from a single well-characterized start site. Mutations within the TdT core promoter demonstrated that the sequence between -3 and +5 functioned as a distinct DNA element that was necessary and sufficient for accurate transcription in the absence of a TATA sequence (Javahery et al., 1994; Smale and Baltimore, 1989). This region of the TdT promoter matched the conserved motif proposed in earlier studies as an Inr element (Corden et al., 1980).

A functional consensus sequence for an Inr was subsequently defined as 5'-(C/T)(C/T)A+iN(T/A)(C/T) (C/T)-3' by examining the ability of initiator mutants and randomly generated sequences to promote transcription in mammalian cells (Javahery et al., 1994; Lo and Smale, 1996). Insertion of the Inr consensus into a synthetic promoter that lacked a TATA box but contained upstream binding sites for the transcription factor Spl was sufficient to support high levels of transcription that initiated from a specific site in the Inr sequence (Smale et al., 1990). Although not all flanking pyrimidine nucleotides (-2, +4 and +5 positions) were essential for Inr function, more activity was associated with motifs that contained more pyrimidines in the flanking positions (Javahery et al., 1994; Lo and Smale, 1996).

In Drosophila, a bioinformatics approach produced the Inr consensus 5'-TCA+1(G/T)T(C/T)-3', a sequence that is very similar but not identical to the functional Inr consensus determined for mammalian cells (Arkhipova, 1995; Hultmark et al., 1986). The A+i position in the Inr is most commonly the site of transcription initiation. However, under conditions transcription does not begin at the A+i nucleotide, the Inr element retained the ability to stimulate the efficiency of transcriptional initiation from the alternative start sites (O'Shea-Greenfield and Smale, 1992). Therefore, the Inr carries out two distinct and separable functions in RNA polymerase II dependent gene transcription.

A: Initiator-binding Proteins

Al: TAF Subunits of TFIID

In DNase I footprinting experiments, TFIID protected sequences downstream of the TATA box including the Inr (Martinez et al., 1994; Parker and Topol, 1984; Purnell and Gilmour, 1993; Sawadogo and Roeder, 1985; Zhou et al., 1992). UV crosslinking and electrophoretic mobility shift assays with recombinant TAF subunits of TFIID revealed that TAF1 and TAF2 directly contacted the Inr (Verrijzer et al., 1995; Verrijzer et al., 1994). Moreover, a TAF1-TAF2 dimeric complex enriched for the Inr consensus in a random double-stranded oligonucleotide site selection screen (Chalkley and Verrijzer, 1999). A temperature-sensitive mutation in TAF1 was also shown to disrupt the ability of TFIID to bind to the Inr of TATA-less promoters (Dehm et al., 2004; Hilton and Wang, 2003). These findings suggest that in the absence of a TATA sequence, the TAF2 and TAF 1 subunits of TFIID recognize the Inr element leading to the stable binding of TFIID to the core promoter.

A2: Transcription Factor //-/

Another model that has been proposed for TFIID recruitment to TATA-less promoters involves Inr binding by a cellular protein, which subsequently interacts with TFIID to bring the transcription factor complex to the core promoter. In support of this model, TFII-I was purified as a protein that binds to the initiator sequence within the adenovirus major late promoter (Roy et al., 1991). The addition of TFII-I to in vitro transcription assays stimulated transcription from a naturally occurring TATA-less, Inr-containing murine V P 5.2 promoter (Cheriyath et al., 1998; Manzano-Winker et al., 1996). The immunodepletion of TFII-I from nuclear extracts with an anti-TFII-I antibody also abolished transcription from the V]3 promoter. These results supported the hypothesis that TFII-I is a general transcription factor involved in initiator recognition. To complicate matters, TFII-I was subsequently found to be identical to proteins that interacted with distal promoter elements in several inducible genes and stabilized the DNA binding of other transcription factors (Grueneberg et al., 1997; Kim et al., 1998; Parker et al., 2001). These findings suggest that TFII-I is a multifunctional transcription factor. Whether the protein acts at the Inr or at more distal control elements may be determined by the cis-elements within the promoter region of interest.

A3: Ying Yang 1

Ying Yang 1 (YY1) is a zinc finger protein that can bind to the Inr of the adeno-associated adenovirus (AAV) core promoter and stimulate Inr dependent transcription in an in vitro transcription system reconstituted with pure proteins (Seto et al., 1991). Mutations that disrupt YY1 binding, however, had no effect when transcription levels were monitored in crude nuclear extracts or in vivo (Javahery et al., 1994; Lo and Smale, 1996). Likewise, Inr mutations that abolished transcriptional activity had little effect on YY1 binding to core promoter DNA. YY1 was originally isolated as a protein that bound to a distal element in the AAV P5 promoter and repressed transcription (Shi et al., 1991). More recently, YY1 has been shown to belong to a family of mammalian polycomb group proteins that function in high molecular weight complexes to repress transcription (Atchison L, 2003). YY1 clearly has an important role in the regulation of gene transcription, but whether its primary site of action is at the Inr remains unclear.

B: Inr Function in TATA-Containing Core promoters

The Inr and TATA box, found together in many core promoters, are binding sites for different components of the TFIID complex. These core promoter elements act synergistically when spaced 25-30 nucleotides apart but function as independent DNA elements when separated by greater than 30 nucleotides (O'Shea-Greenfield and Smale, 1992). The extended footprint reported for TFIID covered approximately 40 bp downstream of the transcription start site, suggesting that the observed synergy could be attributed to the TATA box and Inr cooperatively acting as contact points for the TFIID complex. When these two core promoter elements are greater than 30 nucleotides apart, the binding of a single TFIID molecule would no longer be stabilized by two sites of DNA-protein interaction. Interestingly, if the length of the intervening sequence is reduced to 15 or 20 nucleotides, the TATA box and Inr remained synergistic in supporting transcription (O'Shea-Greenfield and Smale, 1992). However the start site of transcription was now dictated by the position of the TATA box, with transcription initiation occurring 25 bp downstream of the TATA sequence and not within the Inr motif. Therefore, the distance between the TATA box and Inr determines not only how these two core promoter elements functionally interact to influence the level of transcription but also the position of transcription initiation.

Downstream Promoter Element (DPE)

The DPE is a downstream core promoter element located at +28 to +32 relative to the +1 nucleotide of the Inr (Kutach and Kadonaga, 2000). It was originally identified in Drosophila as a DNA recognition site for purified TFIID (Burke and Kadonaga, 1996; Purnell et al., 1994). Using different experimental approaches, a variety of sequences was found to function as a DPE in Drosophila cells. The analysis of natural Drosophila promoters led to the consensus sequence 5'-(A/G)G(A/T) (C/T)GT-3' between +28 and +33, and a preference for a G nucleotide at position +24 (Kutach and Kadonaga, 2000). A comparison of eukaryotic promoters revealed that the DPE is conserved from Drosophila to human and is most commonly, but not exclusively, found in promoters lacking a TATA sequence (Burke and Kadonaga, 1997). However, a homologue of the DPE in Saccharomyces cerevisiae has yet to be discovered.

A: Inr Function in DPE Dependent Transcription

Studies analyzing the Drosophila genome suggest that approximately 8% of Drosophila core promoters contain a DPE (Ohler et al., 2002). Almost 50% of the DPE-containing promoters also contained an Inr while only 11% contained a TATA sequence, indicating a strong bias for the coexistence of the DPE and Inr motifs in core promoters. In TATA-less promoters, TFIID has been shown to contact both the DPE and Inr elements by DNase I footprinting (Burke and Kadonaga, 1996; Purnell et al., 1994). Mutations within either sequence motif severely reduced promoter activity (Burke and Kadonaga, 1996; Burke and Kadonaga, 1997; Kutach and Kadonaga, 2000). The functional relationship between the DPE and Inr is also illustrated by the maintenance of a strict spacing requirement between the two core promoter elements in confirmed DPE-dependent Drosophila promoters (Kutach and Kadonaga, 2000). Removal or insertion of nucleotides between the DPE and Inr led to a significant reduction in TFIID binding and basal transcription levels. These findings suggest that the DPE and Inr cooperate to mediate the interaction of TFIID with core promoter DNA when a binding site for TBP, specifically a TATA box, is absent.

B: TAFs Bind the DPE

TFIID is a multi-subunit complex. Which subunit(s) of TFIID directly makes contact with the DPE? Photo-crosslinking experiments indicated that TAF6 and TAF9 were likely candidates (Burke and Kadonaga, 1997). TAF6 and TAF9 each contain a histone fold domain (HFD), which has been proposed to play a role in direct DNA binding (Gangloff et al., 2001). Experiments with purified proteins revealed that TAF9 but not TAF6 was able to bind to the DPE in a 60 bp promoter fragment from the human IRF-1 gene in electrophoretic mobility shift assays (Shao et al., 2005). A TAF6-TAF9 complex displayed greater DPE binding activity but only when both proteins retained their respective HFDs. The sequence specificity of the heterodimer for DPE binding was also increased compared to the individual TAFs. In competition assays, unlabeled double-stranded oligonucleotides containing wild-type or mutant DPE were equally effective at competing for TAF9 binding to the radio-labeled DPE fragment. With the TAF6-TAF9 complex, wild-type DPE was a significantly better competitor than the mutant DPE sequence. While the HFDs of TAF6 and TAF9 do not directly bind to promoter DNA, the interaction of these domains contributed to the binding affinity and sequence specificity of the TAF-DNA interactions at the DPE.

C: DPE- vs. TATA-Dependent Promoters

The DPE is required for the efficient binding of TFIID to a subset of TATA-less promoter. Therefore, it could be considered a downstream TATA box, as it is serving as a binding site for TFIID. As predicted, the insertion of the DPE at its proper position restored transcriptional activity to promoters that contained a mutant TATA sequence (Burke and Kadonaga, 1996). These findings suggested that transcription at DPE and TATA dependent promoter utilizes similar molecular mechanisms. The identification of NC2 (negative cofactor 2, also known as Drl-Drapl), an activity that stimulated transcription from DPE-dependent promoters but repressed TATA-dependent transcription, indicates otherwise (Willy et al., 2000). In addition, a mutation in NC2 that disrupted activation of DPE-dependent transcription had no effect on the ability of the mutant NC2 to repress TATA-dependent transcription. These results suggest that NC2 carries out distinct and separable functions at DPE- and TATA-driven promoters and that NC2 is able to recognize the fundamental differences between these two types of core promoter elements.

TFIIB Recognition Element (BRE)

The BRE is the fourth core promoter element identified in eukaryotes, after the TATA box, Inr and DPE. The BRE functions as a DNA binding site for the general transcription factor TFIIB (Lagrange et al., 1998). It is found in only a subset of TATA-containing core promoters. The importance of the BRE in transcriptional regulation was first suggested by X-ray crystallography data. The structure of a TBP-TFIIB-TATA box ternary complex showed that TFIIB interacted with the major groove of DNA upstream of the TATA box (Nikolov DB, 1995). In these studies the DNA fragment used to assemble the ternary complex contained only 3 base pairs upstream of the TATA sequence. Subsequent photo-crosslinking experiments confirmed the interaction of TFIIB with core promoter DNA and suggested that the site of contact spanned 7-9 base pairs (Lagrange et al., 1998; Lee and Hahn, 1995). A consensus sequence for the BRE was subsequently determined by binding-site selection using a library of DNA fragments containing 12 randomized nucleotide pairs followed by the TATA element from the adenovirus major late promoter (Lagrange et al., 1998). After two rounds of selection, the 7 bp consensus sequence 5'-(G/C)(G/C)(G/A)CGCC-3' was defined, with the 3' C nucleotide of the BRE immediately followed by the 5' T nucleotide of the TATA motif. The ability of TFIIB to bind to the BRE consensus in the absence of TBP was confirmed by fluorescence anisotropy DNA-binding and site-specific protein-DNA photo-crosslinking. Interestingly, the BRE motif does not appear to be conserved in yeast or plants, suggesting that its function in transcriptional regulation may be restricted to higher eukaryotes.

In vitro transcription studies carried out with purified proteins demonstrated that the BRE stimulated the initiation of transcription by RNA polymerase II. The sequence specific interaction of TFIIB with the core promoter led to the hypothesis that the BRE facilitates the entry of TFIIB into the transcription complex. However, Evans et al later reported a repressor function for BRE when examining basal transcription levels in vitro using crude nuclear extracts and in vivo by transient transfection (Evans et al., 2001). The inhibitory effect of the BRE was overcome by the addition of the transcriptional activator Gal4-VP16 such that comparable levels of activated transcription was detected from wild-type and mutant BRE containing promoters. The net result was that the magnitude of transcriptional activation mediated by DNA bound activator proteins from BRE containing promoters was significantly increased. It will be interest to test if the core promoter content and/or surrounding environment are factors that determine whether the BRE will have a positive or negative role in transcriptional regulation.

Motif Ten Element (MTE)

The most recently defined core promoter element involved in RNA polymerase II transcription is the motif ten element or MTE (Lim et al., 2004). The MTE was recognized as a potential core promoter element when the sequence was found overrepresented near the start site of transcription in nearly 2000 Drosophila genes (Ohler et al., 2002). The identification of known core promoter elements, including the TATA box, Inr, and DPE, at their predicted positions relative to +1, validated the algorithm that led to the discovery of the MTE.

The MTE is conserved from Drosophila to humans and supports transcription by RNA polymerase II in the absence of a TATA motif. MTE activity requires a properly positioned Inr, similar to what has been reported for Inr-DPE dependent transcription. The spacing requirement, however, is more forgiving, as removal of one nucleotide only modestly reduced transcription from Inr-MTE promoters. The identical change in Inr-DPE spacing results in a several-fold decrease in transcriptional activity.

The consensus sequence for the MTE is 5'-C(G/C) A(A/G)C(G/C)(G/C)AACGC-3' and can be found between nucleotides +18 and +29, thereby overlapping the DPE by two nucleotides. Deletion of nucleotides unique to the MTE, in the context of an inactive DPE, compromised transcription levels, with nucleotides +18 to +22 being essential for MTE activity. A search of the Drosophila genome uncovered ten putative MTE dependent core promoters, none of which contained a TATA box. Primer extension analysis using poly(A)+ Drosophila mRNA mapped the start site of transcription to the C.i position of the Inr in 9 out of 10 of the promoters. Using hybrid promoters in which the TATA box and Inr region from the hb?2 core promoter were fused to the downstream promoter region of MTE-containing genes, the MTE restored transcriptional activity to core promoters that contained a TATA box deletion. The MTE could also compensate for the loss of DPE activity. These findings indicate that the MTE is a bonaftde core promoter element that is distinct from the DPE.

A: MTE Binding

Core promoter elements represent points of contact for transcription factors that regulate the process of transcription initiation. The ability of the MTE to compensate for loss of the TATA box or the DPE, both binding sites for TFIID subunits, suggested that the MTE may represent another entry point for TFIID. Mutations within the MTE that disrupted transcriptional activity also reduced the interaction of TFIID with the Inr region, as determined by DNase I protection (Lim et al., 2004). The binding of TFIID to the MTE-containing Tollo promoter was quite weak, suggesting that other factors, yet to be determined, may be contributing to the efficient binding of TFIID to TATA-less MTE-Inr core promoters.

Core Promoter Selectivity of Transcription Regulators

The core promoter is the assembly site for the transcription machinery. However, enhancer-binding proteins ultimately target the core promoter in order to activate transcription. A body of evidence exists indicating that transcriptional enhancers display core promoter selectivity. The ability of such regulatory proteins to activate transcription can be dictated by the DNA elements within a given core promoter. The diversification of core promoters provides an addition level of combinatorial gene regulation.

A: Distinct TATA Box Sequences

The first example of this type of regulation was observed in studies on the his3 promoter in S. cerevisiae. The his3 promoter contains two TATA boxes, named Tc and Tr, each initiating transcription from a distinct site (Struhl, 1986). The downstream Tc, is a canonical TATA box (TATAAAA) whereas the upstream TR is a non-canonical TATA that matches the looser consensus sequence 5'-TATA(A/T)AA(G/A)-3' (Mahadevan and Struhl, 1990). The low level of his3 transcription observed under normal growth conditions is supported by Tc(Iyer and Struhl, 1995). Upon his3 induction, the activator proteins Gal4 and Gcn5 specifically increase transcription from TR but not from Tc. These experiments demonstrated that transcriptional regulators can distinguish between core promoter elements, in the case of the his3 gene, between different TATA sequences.

The ability of enhancer sequences to discriminate between canonical and non-canonical TATA sequences has also been observed in mammalian cells. For example, the myoglobin enhancer is capable of activating transcription from the myoglobin promoter, which contains a canonical TATA sequence, but not from the SV40 early promoter (Wefald et al., 1990). When the SV40 TATA sequence of TATTTAT was changed to the canonical motif, the resulting promoter responded to the myoglobin enhancer. A similar finding was reported for transcription from the hsp70 promoter. Activation of the hsp70 promoter by El A is dictated by the TATA sequence within the core promoter. When the canonical hsp70 TATA element was replaced with the SV40 early promoter TATA sequence, the promoter was no longer activated by E1A (Simon et al., 1988). Both the myoglobin enhancer and the transcriptional activator El A specifically preferred the TATAAAA sequence compared to the TATTTAT sequence within the core promoter for proper function. The ability of enhancers and activator proteins to distinguish between TATA sequences in the core promoter represents one way of targeting their actions to a specific gene.

B: TATA versus DPE Core Promoters

The analysis of gene transcription in Drosophila has revealed that enhancer elements can also distinguish between promoters that are driven by a TATA element versus the DPE. These studies examined the transcriptional activity of enhancers that were located either between or within close proximity of two distinct core promoters. One promoter, from the even-skipped gene, contained a TATA element. The other promoter was the TATA-less, DPE-dependent white promoter. The two enhancers analyzed, AE1 and IAB5, preferentially activated transcription from the TATA containing even-skipped core promoter (Ohtsuki et al., 1998). In the absence of the even-skipped promoter, AE1 and IAB5 activated transcription from the white promoter. Therefore, both enhancer elements possessed the ability to function with the core promoter of the white gene, but, when given a choice, displayed a strong preference for the TATA-containing even-skipped core promoter. Enhancer elements that can discriminate between different core promoter elements is extremely useful when the gene to be expressed is situated within a cluster of other genes or is located many kb away from the enhancer element. Therefore, core promoter elements can play a key role in directing enhancer function to the proper core promoter region.


The identification of core promoter motifs has expanded our understanding of the mechanisms regulating

RNA polymerase II transcription. The TATA box, Inr, DPE, BRE, and MTE have unique and overlapping functions in the process of transcription initiation. However, a significant number of genes do not contain any of these predominant core promoter motifs within their transcriptional regulatory region. Computational genome analysis has revealed additional novel, uncharacterized sequence motifs that are prevalent from -60 to +40 in the core promoter of Drosophila genes (Ohler et al., 2002). Therefore, it is very likely that many more core promoter elements still remain to be discovered and characterized. The long-term goal will be to determine how different combinations of core promoter elements contribute to the regulation of RNA polymerase II dependent transcription in eukaryotic cells.

Yin Yang Balance

Yin Yang Balance

Achieve Health, Wealth And Body Balance Through Yin Yang Mastery. Cut up on the old stone drums of Republic of China, inscribed in books handed down through thousands of years, traced on ancient saucers and on saucers made today, is a sign and a symbol. It is woven into textiles, stitched into embroideries, emblazoned over house gates, wrought into shop emblems, a circle, locked together inside it yang and yin yang, light, yin, dark, each carrying inside itself the essence of the other, each shaped to the other

Get My Free Ebook

Post a comment