A.V. Spirov
The Sechenov Institute of Evolutionary Physiology and Biochemistry,
Russian Academy of Sciences, St. Petersburg, Russia
Discussion: Evolution of Homeodomains and Regulatory Elements
We are not aware of publications on phylogenetic analysis of regulatory regions of any gene families. At present, even principles and directions of evolution of regulatory regions compared to those for coding sequences are unknown. This complicated question might be elucidated to an extent by comparison of results of phylogenetic analysis of translated and nontranslated sequences of homeobox genes.
Our results on phylogenetic analysis of the Antp-like genes compared to the results of other authors allow us to propose the presence of two main tendencies in evolution of regulatory elements of these genes. The first which is predominantly emphasized in the present work is an extremely high conservatism. Analysis of promoter regions of vertebrate genes shows the presence of a great number of conservative sites (including those that were tested experimentally) identical or similar to binding sites of Drosophila transcription factors. There are binding sites of BICOID morphogen (promoter HoxD-4), homeoprotein ENGRAILED (promoter HoxD-7), and products of segmentation-controlling genes hunchback and Kruppel (promoter HoxB-6). Moreover, direct experiments on introduction of gene constructs with parts of human homeobox genes into the Drosophila genome as well as experiments on introduction of elememts of Drosophila genes into the mouse genome, showed essential conservatism not only of individual regulatory elements and homeodomains but also evolutionary conservatism of the systems regulating the activity of homeobox genes as a whole [29-30].
The second tendency is directly opposite. It consists in fast (in the time scale of evolution) changes of target site specificity, disappearance of certain target sites and appearance of others among nonfunctioning spacer sites. These processes definitely are facilitated by similarity of binding sites for different transcription factors. For example, a great number of functionally associated but structurally nonrelated proteins that are products of genes from the cascade of the segmentation control in Drosophila bind with sites differing in one or two nucleotides [31]. Hence, by one or two point mutations of the nucleotide substitution, the binding site may change its specificity.
It should be noted that the present study revealed substantial differences in the structure of regulatory zones in homologous and related vertebrate homeobox genes. The most illustrative examples seem to be regulatory zones of Scr-group genes in which functionally close elements differ qualitatively in their structure. The homeoproteins binding sites and the site for retinoid receptors in the promoters of HoxC-5 genes (Fig. 5) are not homologous to the same sites in the promoters of HoxA-5/HoxB-5 genes (Fig. 4). The second example deals with a set of binding sites in the 300 bp promoter region of HoxD-4 genes and HoxC-4 gene (Fig. 6). There is no target site for HOX-1.3 in HoxC-4 but it is only this gene in the beginning of the promoter zone that has potential sites for binding such universal transcription factors as C/EBP and AP-1.
The examples of the known and proposed regulatory elements in promoter zones of Dfd-, Scr-, and Ubx-group genes summarised in the Table illustrate both conservatism and variability in organisation of regulatory elements of Antp-class genes.
As noted above, HoxA-4, HoxB-4, HoxC-4, and HoxD-4 are the closest evolutionary to the Deformed Drosophila gene which is known to be an example of positively auto-regulatory gene (self-activation by its own protein-product) [32]. Genes of the Dfd-group and, possibly, of Scr-group are quite likely to be characterised by a kind of co-evolution of homeoboxes and promoters. The homeobox is known to code the homeodomain responsible for specific recognition and binding of ATTA-containing sites in regulatory regions of target genes. Thus, the HoxA-5 gene (hox-1.3) product binds specifically with the site in the promoter zone of the HoxA-5 gene (auto-regulation loop), and the same gene product binds specifically with the homologous site in the human HoxB-5 gene (hox-2.1) promoter zone [17]. There are experimental grounds to believe that similar situation occurs in the case of the HCR-domain in the regulatory region of the human HoxD-4 gene, the target for the protein-product of the HoxD-4 gene (hox-4.2), and its HOX-D cluster neighbours (auto- and cross-regulatory action) [18]. Moreover, the protein-product of the HoxA-4 gene (hox-1.4) has been recently shown to bind specifically with the target sites located in the promoter and intron zones of the HoxA-4 gene (an auto-regulation loop) [19]. Therefore, comparative phylogenetic analysis of homeoboxes and auto-regulatory elements of the same group of homologous genes is a unique opportunity to compare evolution of coding and regulatory sequences, the evolution being necessarily coupled.
Comparative analysis of conservative elements in Dfd and Scr families shows the presence of the conservative binding site of the homeoprotein in 9 out of 12 genes presented in the Table. However, analysis of sequences suggests that this element in the HoxC-5 gene group is not homologous to the element in HoxA-5, HoxB-5, and HoxD-4 genes. The sites (including the potential ones) for the action of retinoid receptors demonstrate a similar picture: the pair of direct repeats in HoxB-6 genes most likely is not homologous to the palindrom sequence both in the promoter of HoxA-5 and HoxB-5 genes and in the promoter region of HoxC-5 genes.
box c
______________________________
HoxA-5(mus) cccccattagtGCACGAGTttacctCTAGaggtcaTCAGGCAGGATTTACGACTGGACAACAAA
HoxA-5(hum) CCCCT--TAGTG---G-GTTTA-CT-TA---G------G--AGG-----C--C---A-AACAA-
HoxC-5(hum) ....C--.....---T-AC...-TC-CG---G------A--CAA-----T--G---C-GCTCG-
HoxC-5(mus) ....C--.....---T-AC...-TC-CG---.------A--CAA-----T--G---C-GCTCC-
^^^^^^^^^^^^^^^^^^^
Box B
box d
____________________
HoxA-5(mus) ACAACAAAAGCACGTGATTC.GAA
HoxA-5(hum) A-AACAA-AG-------T--.G--
HoxC-5(hum) C-GCTCG-GT-------C--TA--
HoxC-5(mus) C-GCTCC-GT-------C--TA--
^^^^^^^^^^^^^^^^^^^^^
Box A
Fig. 7. Results of analysis of conservative elements, based on the multiple alignments of the conservative site in the promoter regions of human and mouse HoxC-5 (hox-3.4) and HoxA-5 (hox-1.3) genes.
A nearly complete coincidence of the Box B sequence in the HoxC-5 gene and of the box c sequence in the HoxA-5 gene as well as a good coincidence of Box A in the HoxC-5 gene and of box d in the HoxA-5 gene are reveaked. Target sites of the HoxA-5 genes are marked by lower-case letters (see Fig. 4).
Indeed, on the one hand, the multiple alignments of the promoter zones of the HoxC-5 and HoxA-5 genes reveal location of the HoxC-5 gene target sites outside conservative regions (Fig. 7). On the other hand, the elements of the HoxC-5 genes for the action of retinoid receptors (Box B) as well as for the action of homeoproteins (Box A) are located in conservative regions of multiple sequence alignments (Fig. 7). However, functions of these conservative sequences in the of HoxA-5 gene promoters are unclear.
It should also be noted that the promoter HoxC-4 zone (hox-3.5) differs qualitatively by its sequence from a similar zone of other paralogous genes. This seems to be due to the location of the HOX-3.5 protein homeodomain far from other homeodomains (Fig. 3,B).
Thus, some elements of the gene networks can change qualitatively within the taxon (family, class), whereas other elements of the same network are preserved, in fact, with no changes from lower animals to mammals and humans.
Table. The known and assumed regulatory elements in the promoter region of some vertebrate homeobox genes.
Gene Target site for the Target site for
homeoprotein action the action of nuclear
retinoid receptors
BICOID
HoxB-6(zeb) GAAcggattaAAA ? ATTggtcaGTGGCTggtcaGAT
HoxB-6(mus) .............? ? -----------TTG-A------
ANTP
HoxB-5(mus) atcgtcattaGGT AAttacctCTTGaggtcaT
HoxB-5(zeb) ............. -------------a-----
HOX-1.3
HoxA-5(mus) ACccccattaGTG ..ttacctCTAGaggtcaT
HoxA-5(hum) -----t------- ..-----------------
HOX-1.3
HoxD-4(hum) AActacattaATA GCtcaccAACCCggtgaT
HoxD-4(mus) ----ct----GC- ------------------
HoxC-4(mus) .............? ..................?
ANTP
HoxC-5(hum) ? TCAAGCCAAATTTATGAGG ? CTCgagtcacgtgactcTATT
HoxC-5(mus) ? --------.--------C- ? GGAGCTGCACGTGATTC----
HoxC-5(zeb) ? --------A--------G- ? CTCCAGT-------C------
HoxC-4(mus) .............? ..................?
ANTP
HoxC-5(hum) ? TCAAGCCAAATTTATGAGG ? CTCgagtcacgtgactcTATT
HoxC-5(mus) ? --------.--------C- ? GGAGCTGCACGTGATTC----
HoxC-5(zeb) ? --------A--------G- ? CTCCAGT-------C------
Note: Above the sequence of the binding site of homeoproteins, names of corresponding homeoproteins are indicated; dash - identical nucleotide in the column; points - breaks in aligned sequences; target sites are marked by lower-case letters. hum - man, mus - mouse , zeb - zebrafish
Get your own Free Homepage