This page reports the text and figures of the poster presented by P.-A. Binz at the Second European Meeting of the Protein Society in Cambridge (UK), 1997
A unix compressed postscript file of the poster in A4 format is also available

Note: most of the figures are presented in a reduced size. To see the full size hold the right button of your mouse and choose "view this image"

Molecular Evolution of Metallothioneins:
Contributions from coding and non-coding Regions

P.-A. Binz, J.H.R. Kägi, Institute of Biochemistry, University of Zürich, Winterthurerstr. 190, CH-8057 Zürich, Switzerland

email: pabinz@bioc.unizh.ch, phone 41-(1)-635 65 59; fax: 41-(1)-635 68 05


Introduction

Metallothioneins (MTs) constitute a family of small Cys-rich metalloproteins involved in metal regulation and detoxification processes. Proteinaceous MTs are divided into two classes differing in the relative position of the Cys. Class I subsumes the MTs with the Cys arranged as in the mammalian forms. Class II comprises all forms devoid of such correspondance. Mammals express 4 different subforms, two of them in all tissues (MT-1 and MT-2) and two only in brain (MT-3) and in stratified epithelia (MT-4).

The number of available MT protein sequences is continuously increasing:

81 Vertebrates
31 Invertebrates
12 Microorganisms
46 Plants

170 Total (March 1997)

Vertebrate MTs contain 20 almost totally conserved Cys on a total of 60 to 68 AAs. The other Class I MTs have sequences with less similarity, as illustrated by the alignment of their protein sequences:

Yellow: Cys; blue: AAs identical to the vertebrate consensus sequence; black: AA at position 10 or 11 defining the MT-2 subform; green: AA conserved in the invertebrate sequences.

Although their amino acid sequences are diffferent, mammalian and crustacean MTs are similar in their spatial structures: rat MT-2 (left) and blue crab MT-I (right) are made of two separate tetracoordiated Cd-S clusters surrounded by a polypeptide chain. All Cys (yellow) are involved in the binding of metal (blue). The NMR spectra of the crab MT do not yield the exact relative positions of the two domains which therefore are represented as separate entities.

Results

1. Analysis of the protein sequences

In order to study the evolutionary relationships of MT phylogenetic trees were calculated from the amino acid sequences of vertebrate, crustacean and molluscan protein MT sequences. Different methods were used for this inference. The trees shown were obtained with the maximum parsimony method.

Sequences in bold numbers; bootstrap values (100 replicates) in normal font. Left: % occurence in the most parsimonious trees; the highly variable branching patterns are presented by in the multifurcating nodes I and II. Right: branch lengths in bold; bootstrap values in italics; for both crustacean and molluscan MTs the branching pattern is identical with Fitch topology.

While the data yielded branching patterns resolvable at the protein level in the subfamilies of Class I some uncertainities remained with respect to the relative positions of the monophyletic vertebrate MT subgroups. Also for the mammalian MT-1 and MT-2 subgroups no unambiguous assignment of a single sequential model of evolution for the mammalian subforms was possible. This problem was resolved by extending the analyses to genomic sequences.

2. Analysis of the gene sequences

All vertebrate MT genes display conserved positions of the splicing sites. All mammalian MT genes are located on a single gene cluster:

2.1. Coding segments: the exons

The phylogenetic trees obtained from the exon sequences of the vertebrate genes closely resemble the protein trees. The subgroups are well resolved in the below presented bootstrap maximum parsimony tree. However, even if most of the mammalian MT-1 and MT-2 show tendencies towards two separate subforms, a general separation was not affirmed by the statistical tests.

2.2. Non-coding segments

Introns

Pairwise comparisons of the vertebrate MT gene sequences using dotplots showed subform specific regions of similarity in the intron sequences. As an example, the comparison of human and sheep MT-1 and MT-2 genes highlighted significant similarity in the introns among the two MT-1 and among the two MT-2. No such similarity was found when the two human or the two sheep genes were compared. The same type of observation was made in the gene portion upstream of the first exon.

Coordinate axes correspond to base numbering of the compared genes. The correlated exon sequences are labelled by numbers in the dotplots.

5'untranslated portion (5'UTR) of the genes

The 5'UTR of the mammalian MT-1 and MT-2 mRNAs are stretches of about 80bp. The phylogenetic trees calculated from 22 5'UTR sequences unambigously distinguished the MT-1 and the MT-2 subforms affirming the paralogous character of these two subforms for the mammals.

Maximum parsimony, maximum likelihood and Fitch methods are compared here. U1, R1, U2, R2, P2 refer to ungulate MT-1, rodent MT-1, ungulate MT-2, rodent MT-2 and primate MT-2, respectively. The thick lines highlight branches of high confidence (bootstrap value over 90%). The bootstrap values are given for the separation between the MT-1 and MT-2 subforms.

Untranscribed portion (5'UT) of the genes

This portion of the genes contain numerous regulatory elements. An important one for metallothionein is the so called metal responsive element (MRE). It is generally present in a variable number of copies. Its described consensus sequence is the 15mer nucleotide CTNTGCRCNCGGCCC where R is a purine and N any base and its central highly conserved core sequence TGCRCNC is required for the binding of protein factors. The numbers of MREs and their positions on the coding ("->") or on the non-coding strand ("<-") in 5'UT-region of vertebrate MT genes correlate with subgroup specificity.

Conclusion: proposed model of evolution

All MTs share a number of conspicuous phenotypical features and are thus considered as constituting a protein superfamily. The alignment of their AA sequences does not allow, however, to unite them in a single pedigree of evolutional divergence. Phylogenetic relationships are, however, discernable in the structure of MTs from different taxonomic classes of animals and plants. Thus the evolution of the vertebrate sequences can be represented as originating from a single vertebrate MT ancestor gene (avMT):

The difffereces in the subform manifested in the 5'UTR and the 5'UT segments may be related to observed specific differences in the mechanism of transcription and translation of the corresponding genes.

Final remarks

This work was part of my PhD thesis at the Institute of Biochemistry of the University of Zürich.

For more informations on metallothionein, read for instance Meth. Enz 205: Metallothionein and Related Molecules; Academic Press 1991. For more informations about phylogenetic methods, read for instance Felsenstein J. (1988) Phylogenies from molecular sequences: inference and reliability. Annu. Rev. Genet. 22:521-565.

Most of the computer work was performed on Silicon Graphics workstations, using the Wisconsin GCG Software Package and the PHYLIP package of Joe Felsenstein (available via "ftp" at "ftp.genetics.washington.edu").

Pierre-Alain Binz, April 1997, text revised October 1997, page updated February 20th, 2001