Abstract

Searches of several sequence databases reveal that human HD1, yeast HDA1, yeast RPD3 and other eukaryotic histone deacetylases share nine motifs with archaeal and eubacterial enzymes, including acetoin utilization protein (acuC) and acetylpolyamine amidohydrolase. Histone deacetylase and acetylpolyamine amidohydrolase also share profound functional similarities in that both: (i) recognize an acetylated aminoalkyl group; (ii) catalyze the removal of the acetyl group by cleaving an amide bond; (iii) increase the positive charge of the substrate. Stabilization of nucleosomal DNA-histone interaction brought about by the change in charge has been implicated as the underlying cause for histone deacetylase-mediated transcriptional repression. We speculate that the eukaryotic histone deacetylases originated from a prokaryotic enzyme similar to the acetylpolyamine amidohydrolases that relied on reversible acetylation and deacetylation of the aminoalkyl group of a DNA binding molecule to achieve a gene regulatory effect.

Introduction

Histones and DNA constitute the nucleosomes which are essential in packaging eukaryotic DNA into chromosomes. In addition to their structural function, nucleosomes have a role in regulation of transcription that is, at least in part, attributed to reversible acetylation and deacetylation of lysine residues clustered near the hydrophilic N-terminal ends of the core histones (for a review see 1–3). Histone deacetylases catalyze removal of the acetyl group from the ε-amino group of lysine side chains, thereby reconstituting its positive charge. It has been suggested that this, in turn, stabilizes interaction between histones and DNA (4) and might make promoter elements inaccessible to activating transcription factors. Accordingly, deacetylation (or the elimination of histone acetylation) is often correlated with transcriptional silencing (see for example 5) and histones in heterochromatin regions are generally hypoacetylated. Conversely, histone hyperacetylation leads to destabilization of internucleosomal contacts and increases accessibility of internucleosomal DNA to transcription factors such as TFIIIA (6) and GAL4-AH (7). This model is certainly an oversimplification and the fact that the histone deacetylases in Drosophila melanogaster and Saccharomyces cerevisiae can also counteract genomic silencing in genes that are located adjacent to telomeres or centromeric heterochromatin (8,9) shows that the entire range of histone-mediated gene regulatory mechanisms is more complex.

While much is known about histone acetyltransferases (reviewed in 2,10,11), histone deacetylases are only now being studied in greater detail. Recently the catalytic subunit of a heterodimeric protein with histone deacetylase activity was purified by affinity chromatography and its gene (HD1) sequenced (12). Two potential histone deacetylases (human and mouse RPD3) have been shown to enhance transcriptional repression activity of the mammalian zinc finger transcription factor YY1 (13) and yeast proteins HDA1 and RPD3 have been identified as the catalytic subunits of two distinct histone deacetylase complexes, HDA and HDB respectively (9,14). Moreover, sequence similarity between the histone deacetylases and a prokaryotic enzyme has been noted (9). To elucidate the functional and phyletic relationships between the histone deacetylases and their bacterial counterparts, we have searched the sequence databases for molecules with similarity to HD1. The present communication shows that histone deacetylases share several blocks of significant sequence similarity with two groups of mostly prokaryotic enzymes, the acuC gene products and the acetylpolyamine amidohydrolases. Our findings suggest that gene regulatory mechanisms that rely on reversible acetylation and deacetylation of DNA binding cations, like lysine side chain groups or polyamines, might have predated the origin of histones.

Materials and Methods

Computational methods

We used BLASTP, TBLASTN (scores >70, P values <0.05) (15) and MOST (r = 0.05) (16) to search the non-redundant peptide and nucleotide sequence database at the National Center for Biotechnology Information (February 20, 1997) for proteins and open reading frames (ORFs) similar to human histone deacetylase HD1 (U50079) (12). A multiple alignment has been produced using PROBE, a new multiple alignment and database search procedure (17). Phylogenetic analyses have been conducted using maximum parsimony (PAUP 3.1) (18) and distance-based methods, protdist/neighbor joining from the PHYLIP software package (19,20).

Figure 1

Multiple sequence alignment of the regions of significant similarity in histone deacetylases, acuC gene product and acetylpolyamine amidohydrolases. Residues that are identical to the human HD1 protein (1277084) are white on a black background. Numbers between dashes indicate the number of residues that have been left out to improve the alignment. Asterisks indicate the putative zinc binding site of the M.ramosa acetylpolyamine amidohydrolase (22); it should be noted though that except for His137, these are not residues that are conserved between these proteins. The alignment has been formatted using ALSCRIPT software (31). The following sequences are included in the alignment: HD1 (human), histone deacetylase HD1 from H.sapiens (U50079/1277084); RPD3 (human), H.sapiens RPD3 (U31814/1667394); RPD3 (mouse), M.musculus RPD3 (U31758/1667396); histone deacetyl. (frog), X.laevis ORF (X78454/602098); histone deacetyl. (fly), D.melanogaster ORF (Y09258/1666637); C08B11.2 (C.elegans), hypothetical 57.1 kDa protein from cosmid C08B11 of C.elegans chromosome 11 (Z46676/1176665); F41H10.6 (C.elegans), ORF from cosmid F41H10 from chromosome 3 or 4 of C.elegans strain Bristol N2 (U61954/1397334); RPD3 (yeast), transcriptional regulatory protein of chromosome 14 of S.cerevisiae (P32561/417699); HOS2 (yeast), ORF from S.cerevisiae chromosome VII (X91837/1177634); acuC (S.xylosus), acetoin utilization protein from S.xylosus (X95439/1177686); acuC (B.subtilis), acetoin utilization protein of B.subtilis (P39067/728801); hypothet. (Syne PCC7002), hypothetical 34.1 kDa protein of Synechococcus PCC7002, (P28606/140712); APAH (M.ramosa), acetylpolyamine amidohydrolase of M.ramosa (D10463/520593); APAH (Sy_cy. PCC6803), acetylpolyamine amidohydrolase ORF from the cyanobacterium Synechocystis PCC6803, (D90900/1651782); APAH (M.jannaschii), acetylpolyamine amidohydrolase ORF of M.jannaschii, (U67502/1591238). Eight sequences that scored significantly in the BLAST searches have not been included in the multiple alignment (nucleotide and protein identifiers in parentheses): a human histone deacetylase ORF (D50405/1665723) with very few differences from HD1; a 201 residue fragment of the RPD3 gene from S.cerevisiae (Z46259/854536); a yeast ORF (X91837/1177634) identical to ORF YGL194c (HOS2) from S.cerevisiae except for a stretch of 12 amino acids between positions 352 and 364; two fragments (Z13965/669044 and Z13965/669045) of the Synechococcus PCC7002 ORF; a fragment (S43160/542653) with sequence identity to the full-length X.laevis histone deacetylase ORF; an untranslated ORF with one or more frameshift errors from the beta proteobacterium Alcaligenes eutrophus (X85729/no CDS feature) with a sequence most similar to the hypothetical 34.1 kDa protein of Synechococcus PCC7002.

Figure 1

Multiple sequence alignment of the regions of significant similarity in histone deacetylases, acuC gene product and acetylpolyamine amidohydrolases. Residues that are identical to the human HD1 protein (1277084) are white on a black background. Numbers between dashes indicate the number of residues that have been left out to improve the alignment. Asterisks indicate the putative zinc binding site of the M.ramosa acetylpolyamine amidohydrolase (22); it should be noted though that except for His137, these are not residues that are conserved between these proteins. The alignment has been formatted using ALSCRIPT software (31). The following sequences are included in the alignment: HD1 (human), histone deacetylase HD1 from H.sapiens (U50079/1277084); RPD3 (human), H.sapiens RPD3 (U31814/1667394); RPD3 (mouse), M.musculus RPD3 (U31758/1667396); histone deacetyl. (frog), X.laevis ORF (X78454/602098); histone deacetyl. (fly), D.melanogaster ORF (Y09258/1666637); C08B11.2 (C.elegans), hypothetical 57.1 kDa protein from cosmid C08B11 of C.elegans chromosome 11 (Z46676/1176665); F41H10.6 (C.elegans), ORF from cosmid F41H10 from chromosome 3 or 4 of C.elegans strain Bristol N2 (U61954/1397334); RPD3 (yeast), transcriptional regulatory protein of chromosome 14 of S.cerevisiae (P32561/417699); HOS2 (yeast), ORF from S.cerevisiae chromosome VII (X91837/1177634); acuC (S.xylosus), acetoin utilization protein from S.xylosus (X95439/1177686); acuC (B.subtilis), acetoin utilization protein of B.subtilis (P39067/728801); hypothet. (Syne PCC7002), hypothetical 34.1 kDa protein of Synechococcus PCC7002, (P28606/140712); APAH (M.ramosa), acetylpolyamine amidohydrolase of M.ramosa (D10463/520593); APAH (Sy_cy. PCC6803), acetylpolyamine amidohydrolase ORF from the cyanobacterium Synechocystis PCC6803, (D90900/1651782); APAH (M.jannaschii), acetylpolyamine amidohydrolase ORF of M.jannaschii, (U67502/1591238). Eight sequences that scored significantly in the BLAST searches have not been included in the multiple alignment (nucleotide and protein identifiers in parentheses): a human histone deacetylase ORF (D50405/1665723) with very few differences from HD1; a 201 residue fragment of the RPD3 gene from S.cerevisiae (Z46259/854536); a yeast ORF (X91837/1177634) identical to ORF YGL194c (HOS2) from S.cerevisiae except for a stretch of 12 amino acids between positions 352 and 364; two fragments (Z13965/669044 and Z13965/669045) of the Synechococcus PCC7002 ORF; a fragment (S43160/542653) with sequence identity to the full-length X.laevis histone deacetylase ORF; an untranslated ORF with one or more frameshift errors from the beta proteobacterium Alcaligenes eutrophus (X85729/no CDS feature) with a sequence most similar to the hypothetical 34.1 kDa protein of Synechococcus PCC7002.

Results and Discussion

Three groups of sequences show similarity to HD1

Our searches identify three groups of proteins (and ORFs) that share statistically significant sequence similarity with HD1, including: (i) a group of exclusively eukaryotic proteins; (ii) the acuC gene product of Eubacteria; (iii) a group of proteins that contain eukaryotic histone deacetylases and prokaryotic acetylpolyamine amidohydrolases (Figs 1 and 2). The alignment comprises 163 residues distributed among nine sequence blocks. It includes a region of ∼70 amino acids with almost no length variation between alignment positions 77 and 136 with many highly conserved and 11 invariant residues (Fig. 1).

Interestingly, 12 of the 20 invariant residues in our alignment are either histidine (6), cysteine (1), aspartic acid (1) or glutamic acid (4). It seems, therefore, conceivable that several of these amino acids are conserved in these proteins across all domains of life because they are involved in binding a metal atom. Indeed, one of the proteins included in our alignment (the acetylpolyamine amidohydrolase of Mycoplana ramosa) is a homodimer that is known to contain one zinc atom per subunit (21,22). So far, it is not known whether histone deacetylases or the acuC gene product also contain bound metal atoms.

Figure 2

Phylogenetic inferences for histone deacetylase and similar proteins. Phylogenetic analysis of the sequence positions used in the multiple alignment in Figure 1. Bootstrap methods provided estimates of support for topological elements (32). Bootstrap values were recorded independently for 1000 resamplings of maximum parsimony (first figure) and neighbor joining (second figure) analysis and branches that did not have at least 75% bootstrap support in one of the two methods have been collapsed to yield polytomies. For both methods we avoided potential bias of input order by employing randomized orders of taxon addition. The tree is unrooted. Sequences that appear in the alignment in Figure 1 are marked with an asterisk (*); additional sequences and their nucleotide and protein identification numbers (where applicable) are as follows: F43G6.4, ORF from C.elegans cosmid F43G6 (Z50070/1066965); URF (1478271), untranslated ORF from Schizosaccharomyces pombe cosmid 800 (U41410/no CDS feature); HDA1, catalytic subunit of the S.cerevisiae histone deacetylase-A (HDA) complex (Z71297/1730711); HOS3, ORF from cosmid 8209/8002 on chromosome 16 of S.cerevisiae (U43503/1163098); histone deacetylase (mouse), ORF from M.musculus (X98207/1771286); URF (J03579), untranslated ORF with two frameshift errors from a chicken mRNA the 54-half of which codes for protooncogenic tyrosine kinase (c-tkl) (J03579/no CDS feature); HOS1, ORF from S.cerevisiae strain AB972 chromosome 16 (Z49219/1084790).

Figure 2

Phylogenetic inferences for histone deacetylase and similar proteins. Phylogenetic analysis of the sequence positions used in the multiple alignment in Figure 1. Bootstrap methods provided estimates of support for topological elements (32). Bootstrap values were recorded independently for 1000 resamplings of maximum parsimony (first figure) and neighbor joining (second figure) analysis and branches that did not have at least 75% bootstrap support in one of the two methods have been collapsed to yield polytomies. For both methods we avoided potential bias of input order by employing randomized orders of taxon addition. The tree is unrooted. Sequences that appear in the alignment in Figure 1 are marked with an asterisk (*); additional sequences and their nucleotide and protein identification numbers (where applicable) are as follows: F43G6.4, ORF from C.elegans cosmid F43G6 (Z50070/1066965); URF (1478271), untranslated ORF from Schizosaccharomyces pombe cosmid 800 (U41410/no CDS feature); HDA1, catalytic subunit of the S.cerevisiae histone deacetylase-A (HDA) complex (Z71297/1730711); HOS3, ORF from cosmid 8209/8002 on chromosome 16 of S.cerevisiae (U43503/1163098); histone deacetylase (mouse), ORF from M.musculus (X98207/1771286); URF (J03579), untranslated ORF with two frameshift errors from a chicken mRNA the 54-half of which codes for protooncogenic tyrosine kinase (c-tkl) (J03579/no CDS feature); HOS1, ORF from S.cerevisiae strain AB972 chromosome 16 (Z49219/1084790).

Histone deacetylases

The highest similarity to HD1 can be found in a group of exclusively eukaryotic proteins that were originally identified because of their similarity to the yeast transcriptional regulatory protein RPD3 (23). This group includes sequences from Homo sapiens, Mus musculus, Gallus gallus (chicken), Xenopus laevis (frog), D.melanogaster, Caenorhabditis elegans and S.cerevisiae. Histone deacetylases have been named in a largely random manner and are identified as either RPD3, HOS2, HD1 or histone deacetylase regardless of specific sequence similarities. While human and mouse RPD3 are similar enough to conceivably be orthologs, they are much more similar to HD1 than to, for example, yeast RPD3 (see Fig. 2). The existence of several histone deacetylases in human and mouse suggests that there was at least one gene duplication event that predated the divergence of the rodent and the primate lineages. As there are at least five putative yeast histone deacetylases (RPD3, HDA1, HOS1, HOS2 and HOS3) (9) and three C.elegans ORFs, it is likely that histone deacetylases form a multigene family with many more members still to be discovered.

Experimental evidence shows that HD1 has histone deacetylase activity (12) and that binding to RPD3 enhances the transcriptional repression activity of mammalian zinc finger transcription factor YY1 (13). Given the high sequence similarity between human HD1 and RPD3 (85% identical), it seems conceivable that histone deacetylases have at least two domains: one that provides the deacetylating chemistry and another that directs the deacetylating effect towards specific genes by interaction with YY1 and other transcription factors (see also 24). Furthermore, we hypothesize that the enzymes in this group, including yeast HOS1 and HOS2 and the C.elegans ORF C08B11.2 (see Fig. 2), are bona fide histone deacetylases with a mechanism of action similar to that described above.

Acetylpolyamine amidohydrolase

The second group of proteins with similarities to human histone deacetylase HD1 includes a group of enzymes referred to as acetylpolyamine amidohydrolase, deacetylase or acetylhydrolase from the alpha proteobacterium M.ramosa, the cyanobacterium Synechocystis PCC6803, the archaeon Methanococcus jannaschii and a number of eukaryotic organisms (Fig. 2).

Polyamines (putrescine, spermidine, spermine, etc.) are widely distributed among both eukaryotic and prokaryotic cells. Invariably they are found complexed with nucleic acids, both DNA and RNA, and their metabolites are essential for normal cell growth (25). However, details of their function have not been fully elucidated (25,26). Acetylpolyamine amidohydrolases catalyze deacetylation of a polyamine by cleaving a non-peptide amide bond, a reaction with obvious parallels to histone deacetylation (Fig. 3). There are a number of different acetylpolyamine amidohydrolases, some of which specifically deacetylate acetylputrescine (e.g. in Streptomyces avellaneus, Arthrobacter spp. and Micrococcus rubes), while the M.ramosa enzyme catalyzes deacetylation of several acetylpolyamines (22). The acetylpolyamine amidohydrolase of M.ramosa is composed of two identical subunits, each of which contains one zinc atom (21,22), but it is not known whether the zinc atoms are involved in the catalytic reaction (21).

Surprisingly, some of the eukaryotic histone deacetylases (e.g. HDA1) are more similar to the amidohydrolases than to the group of histone deacetylases that includes RPD3 and HD1 (Fig. 2). While the yeast HDA1 enzyme has been shown to be a histone deacetylase, this remains uncertain for the other eukaryotic enzymes in this group. As acetylspermidine amidohydrolases do occur in eukaryotes (27), experimental evidence would be required to find out in every single case whether the eukaryotic enzymes in this group deacetylate histones or polyamines, or both. Given the similarities in sequence and chemistry (Fig. 3), it seems conceivable that the histone deacetylases have evolved more than once from polyamine deacetylases and that the products of some of the more recent gene duplication events may have retained some of their (ancestral) acetylpolyamine deacetylating activity.

Figure 3

Reactions catalyzed by the acetylpolyamine amidohydrolase and histone deacetylase. At the top is shown conversion from acetylspermidine to spermidine. This reaction is catalyzed by acetylspermidine acetylhydrolase (also called amidohydrolase or deacetylase). The deacetylation of acetylated lysine residues is catalyzed by histone deacetylase (bottom). Both enzymes recognize the same target (an acetylaminopropyl group) and catalyze the removal of an acetyl group by cleaving a non-peptide amide bond.

Figure 3

Reactions catalyzed by the acetylpolyamine amidohydrolase and histone deacetylase. At the top is shown conversion from acetylspermidine to spermidine. This reaction is catalyzed by acetylspermidine acetylhydrolase (also called amidohydrolase or deacetylase). The deacetylation of acetylated lysine residues is catalyzed by histone deacetylase (bottom). Both enzymes recognize the same target (an acetylaminopropyl group) and catalyze the removal of an acetyl group by cleaving a non-peptide amide bond.

The acuC gene product

A third group of sequences with significant similarity to human histone deacetylase HD1 comprises the gene for the eubacterial acetoin utilization protein (acuC) from Bacillus subtilis and Staphylococcus xylosus (Fig. 1). The acuC gene is part of the acuABC operon, which is found in close proximity to the acetyl-CoA synthetase gene (28).

While the reaction catalyzed by the acuC protein is not known, in B.subtilis it has been shown experimentally that disruption of the acuABC operon by insertional inactivation into the acuA gene results in greatly diminished growth on acetoin and butanediol (28). Acetoin (CH3-CO-HCOH-CH3) is a fermentation product of many bacteria that can be converted to acetate via the butanediol cycle when other carbon sources are depleted (29). As the degradation of acetoin involves deacetylation of acetoin or other intermediates of the butanediol cycle, it is conceivable that the acuC gene product shares deacetylating chemistry with histone deacetylase and polyamine amidohydrolase. However, as opposed to the amide cleaving chemistry of the latter two enzymes, the reaction catalyzed by the acuC gene product would likely involve breaking a carbon-carbon bond. In addition, it has been noted that the first gene in the operon (acuA) contains a motif that is conserved in a number of acetyltransferases (28,30), which also seems compatible with involvement of the acuC gene product in a deacetylation reaction.

Given the uncertainties about the exact role of the acuC gene product, it seems worth re-investigating whether the diminished acetoin utilization in B.subtilis is directly caused by inactivation of a metabolic enzyme or whether acuC has a gene regulatory function and chemistry similar to acetylpolyamine amidohydrolases and histone deacetylases.

Conclusions

Acetylpolyamine amidohydrolases and histone deacetylases are members of a deacetylase superfamily of proteins that share nine blocks of significant sequence similarity. Acetylpolyamine amidohydrolases and histone deacetylases recognize the same target (an acetylaminopropyl group) and catalyze the removal of an acetyl group by cleaving a non-peptide amide bond. Moreover, the activities of both enzymes have the same effect in that they reconstitute the positive charge of the target molecule (polyamines and lysine residues respectively) and thereby increase its affinity for negatively charged molecules (e.g. DNA). If DNA binding by polyamines is generally comparable with DNA binding by histones, then acetylation/deacetylation of polyamines could also result in increased/decreased accessibility of polyamine-bound DNA to DNA binding proteins. We speculate that before eukaryotes and histones evolved there was a gene regulatory mechanism in place that relied on reversible acetylation and deacetylation of DNA binding cations to influence DNA organization. If so, once histones evolved, duplication of one (or several) of the acetylpolyamine amidohydrolase genes could have provided proteins that needed to undergo only a slight shift in substrate specificity to function as histone deacetylases.

Acknowledgements

We thank Andy Neuwald for allowing us to use his PROBE program prior to release and Andy Baxevanis for help with the ALSCRIPT software. We thank Andy Neuwald and Eugene Koonin for critically reviewing the manuscript.

References

1
van Holde
K.E.
Chromatin
 , 
1989
Berlin, Germany
Springer-Verlag
2
Sternglanz
R.
Trends Biochem. Sci.
 , 
1996
, vol. 
21
 (pg. 
357
-
358
)
3
Wolffe
A.P.
Pruss
D.
Cell
 , 
1996
, vol. 
84
 (pg. 
817
-
819
)
4
Allfrey
V.G.
Faulkner
R.
Mirsky
A.E.
Proc. Natl. Acad. Sci. USA
 , 
1964
, vol. 
51
 (pg. 
786
-
794
)
5
Braunstein
M.
Rose
A.B.
Holmes
S.G.
Allis
C.D.
Broach
J.R.
Genes Dev.
 , 
1993
, vol. 
7
 (pg. 
592
-
604
)
6
Lee
D.Y.
Hayes
J.J.
Pruss
D.
Wolffe
A.P.
Cell
 , 
1993
, vol. 
72
 (pg. 
73
-
84
)
7
Vettese-Dadey
M.
Walter
P.
Chen
H.
Juan
L.J.
Workman
J.L.
Mol. Cell. Biol.
 , 
1994
, vol. 
14
 (pg. 
970
-
981
)
8
De Rubertis
F.
Kadosh
D.
Henchoz
S.
Pauli
D.
Reuter
G.
Struhl
K.
Spierer
P.
Nature
 , 
1996
, vol. 
384
 (pg. 
589
-
591
)
9
Rundlett
S.E.
Carmen
A.A.
Kobayashi
R.
Bavykin
S.
Turner
B.M.
Grunstein
M.
Proc. Natl. Acad. Sci. USA
 , 
1996
, vol. 
93
 (pg. 
14503
-
14508
)
10
Roth
S.Y.
Nature Genet.
 , 
1996
, vol. 
14
 (pg. 
3
-
4
)
11
Pennisi
E.
Science
 , 
1997
, vol. 
275
 (pg. 
155
-
157
)
12
Taunton
J.
Hassig
C.A.
Schreiber
S.L.
Science
 , 
1996
, vol. 
272
 (pg. 
408
-
411
)
13
Yang
W.M.
Inouye
C.J.
Zeng
Y.
Bearss
D.
Seto
E.
Proc. Natl. Acad. Sci. USA
 , 
1996
, vol. 
93
 (pg. 
12845
-
12850
)
14
Carmen
A.A.
Rundlett
S.E.
Grunstein
M.
J. Biol. Chem.
 , 
1996
, vol. 
271
 (pg. 
15837
-
15844
)
15
Altschul
S.F.
Gish
W.
Miller
W.
Myers
E.W.
Lipman
D.J.
J. Mol. Biol.
 , 
1990
, vol. 
215
 (pg. 
403
-
10
)
16
Tatusov
R.L.
Koonin
E.V.
Comput. Appl. Biosci.
 , 
1994
, vol. 
10
 (pg. 
457
-
459
)
17
Neuwald
A.F.
Liu
J.S.
Lipman
D.J.
Lawrence
C.E.
Nucleic Acids Res.
 , 
1997
, vol. 
25
 (pg. 
1665
-
77
)
18
Swofford
D.L.
PAUP 3.1
 , 
1990
Champaign, IL
Illinois Natural History Survey
19
Felsenstein
J.
Cladistics
 , 
1989
, vol. 
5
 (pg. 
164
-
166
)
20
Felsenstein
J.
The PHYLIP Software Package v3.5c
 , 
1993
Seattle, WA
Department of Genetics, University of Washington
21
Fujishiro
K.
Ando
M.
Uwajima
T.
Biochem. Biophys. Res. Commun.
 , 
1988
, vol. 
157
 (pg. 
1169
-
1174
)
22
Sakurada
K.
Ohta
T.
Fujishiro
K.
Hasegawa
M.
Aisaka
K.
J. Bacteriol.
 , 
1996
, vol. 
178
 (pg. 
5781
-
5786
)
23
Vidal
M.
Gaber
R.F.
Mol. Cell. Biol.
 , 
1991
, vol. 
11
 (pg. 
6317
-
6327
)
24
Wolffe
A.P.
Science
 , 
1996
, vol. 
272
 (pg. 
371
-
372
)
25
Tabor
C.W.
Tabor
H.
Annu. Rev. Biochem.
 , 
1984
, vol. 
53
 (pg. 
749
-
790
)
26
Marton
L.J.
Pegg
A.E.
Annu. Rev. Pharmacol. Toxicol.
 , 
1995
, vol. 
35
 (pg. 
55
-
91
)
27
Seiler
N.
Can. J. Physiol. Pharmacol.
 , 
1987
, vol. 
65
 (pg. 
2024
-
2035
)
28
Grundy
F.J.
Waters
D.A.
Takova
T.Y.
Henkin
T.M.
Mol. Microbiol.
 , 
1993
, vol. 
10
 (pg. 
259
-
271
)
29
Lopez
J.
Fortnagel
P.
Biochim. Biophys. Acta
 , 
1972
, vol. 
279
 (pg. 
554
-
560
)
30
Neuwald
A.F.
Landsman
D.
Trends Biochem. Sci.
 , 
1997
, vol. 
22
 (pg. 
154
-
5
)
31
Barton
G.J.
Protein Engng
 , 
1993
, vol. 
6
 (pg. 
37
-
40
)
32
Felsenstein
J.
Evolution
 , 
1985
, vol. 
39
 (pg. 
783
-
791
)

Comments

0 Comments