1998 Oxford University Press
21 47914796
SURVEY AND SUMMARY
Natural covalent complexes of nucleic acids and
proteins: some comments on practice and theory
on the path from well-known complexes to new ones
Yu. F. Drygin*
A.N.Belozersky Institute of Physico-Chemical Biology, Moscow State University, Moscow 1198999, Russia
Received March 3, 1998; Revised June 30, 1998; Accepted July 29, 1998
INTRODUCTION
More than 20 years have passed since the discovery of natural
covalent complexes of nucleic acids and proteins (15). Covalent
complexes of nucleic acids and proteins are both stable and highly
specific because they are nucleoproteins linked by a unique
covalent bond. Two important properties are associated with
these complexes: (i) a novel structure, describing a new class of
natural nucleoproteins; (ii) specific functions, that have led to an
understanding of previously undiscovered mechanisms of DNA
and RNA replication (i.e. the protein priming mechanism) and
that are ultimately needed to change the topology of DNA, for
DNA integration, conjugation, DNA structure resolution and
other as yet unknown functions. Due to availability and the ability
to obtain homogeneous preparations, the majority of data on
structure and function have been obtained with covalent nucleic
acid and protein complexes of viral origin.
HISTORY
Covalent binding of polypeptides to nucleic acids was originally
discovered following isolation of high molecular weight viral,
bacterial or eukaryotic nucleic acids. Trace amounts (0.10.3%
by mass) of material composed of amino acids were found in
nucleic acid preparations even after protease treatment and
SDSphenol extraction. Thus it is estimated that as many as 100
polypeptide moieties might be bound to one molecule of
eukaryotic chromosomal DNA via an internucleotide phosphate
linkage. Based on the stability of isolated nucleotide-peptides in
alkaline solution, polypeptide binding to internucleotide phosphate
groups via a phosphoamide bond was proposed (6,7) more than
20 years ago. At that time, no data on the alkali stability of
tyrosine and nucleotide phosphoesters were available.
Subsequently, the phosphotriester character of DNApeptide
covalent complexes was demonstrated (8). This kind of interpoly-
meric linkage has to attenuate adjacent internucleotide bonds and
might lead to an apparent molecular mass decrease during DNA
isolation.
The discovery of phosphoserine in DNA hydrolysis products
(9,10) and the correlation between content of residual peptides
*Tel: +7 095 939 5529; Fax: +7 095 939 3181; Email: drygin@libro.genebee.msu.su
Nucleic Acids Research, 1998, Vol. 26, No.
and proliferation of the cell (11) generated much speculation on
the functions of such complexes. For example, a structure of
chromosomal DNA with subunits of DNA linked to each other
via peptide bridges was proposed (9). This model of bacterial
chromosomal DNA organization was examined and refuted (7).
The initial findings that suggested the existence of nucleic
acidprotein complexes also spurred the synthesis and study of
hydrolytic properties of a variety of model compounds (1215).
The shortcomings of these original studies of natural complexes
included lack of a complete chemical structure of the nucleic
acidprotein covalent linkage unit and absence of the molecular
and functional characteristics of the protein constituents. These
deficiencies were partially overcome by J.Wangs attractive
proposal to explain how topoisomerase (Topo) could change the
topology of supercoiled DNAs. He hypothesized the introduction
of a nick in the DNA helix by a nicking-closing enzyme and
formation of a transient covalent DNATopo I complex active in
restoration of the internucleotide bond (16). Later, formation and
disjoining of the active covalent compound of X174 phage DNA
and A protein during initiationtermination of DNA replication
and segregation of parent and progeny genomes was illustrated in
detail (17). To date, a covalent character of the interpolymeric
bond has been proposed for a large number of viral nucleic
acidprotein complexes (reviewed in 1822). In many cases,
these proposals have been confirmed.
The latest progress in structural analysis of these covalent
complexes was obtained by X-ray diffraction study of DNAprotein
crystals, as well as refinement of the chemical mechanisms
leading to covalent complex formation (23,24).
ISOLATION AND STRUCTURAL ANALYSIS OF
NUCLEIC ACIDPROTEIN COMPLEXES TO
DEMONSTRATE THE COVALENT NATURE
OF THE LINKAGE
We have determined the chemical structure of the interpolymeric
covalent bonds of five nucleic acidprotein covalent complexes
from bacteriophage (25,26), viruses (2729) and cells (30). To
unequivocally prove covalent binding of the protein to the nucleic
4792 Nucleic Acids Research, 1998, Vol. 26, No. 21
acid in the complex, biochemical structural analysis is required.
Some practical comments on these structural studies are given
below.
Resistance to chaotropic agent treatment does not adequately
prove the covalent nature of an interpolymeric bond formed in a
high molecular weight, naturally occurring complex. There are
reports that some cellular non-covalent nucleic acidprotein
complexes tolerate proteinase K + SDS + phenol treatment (8).
The complex formed by biotin and avidin is stable even at high
pH and on exposure to protein denaturing agents, proteolytic
enzymes and organic solvents. Phenol + SDS treatment seems to
be the most powerful method for removing proteins from
nucleoproteins. However, it is important to remember that both of
these one-of-a-kind constituents of the complex protect each
other from dissociation and enzyme digestion and that the
distribution of proteins during phenol treatment of nucleoproteins
depends on the coefficient of extraction (distribution) of the
individual protein. For example, the strongly acidic protein
prothymosin partitions to the aqueous phase quantitatively
during phenol extraction (31). Some bacterial glycopeptide
structures also remain in the aqueous phase after SDS + phenol
extraction (32).
Natural covalent complexes of nucleic acids and proteins are often
available only in picomolar quantities. Thus radioactive labelling of
the complex under study to high specific activities is necessary. All
covalent complexes examined so far are phosphoesters of nucleic
acids and proteins. To label the nucleic acid in the complex in
vivo, radioactive orthophosphate is favoured, because both
constituents share a common phosphate group which occupies a
central position in the linkage unit. It is significant that the specific
radioactivity of the complex must be >200 c.p.m./nucleotide
phosphate to identify the O-phosphoamino acid residue after acid
hydrolysis. To achieve the appropriate level of incorporation of
32P into the complexes, mammalian cells are cultivated in a low
phosphate medium (usually <25 M) in the presence of up to
250 Ci/ml radioactive phosphate. Labelling of the covalent
linkage unit in a complex with a 3H- or 14C-labelled nucleoside
is also helpful. It should be mentioned that RNA labelling in vivo
with [6-3H]uridine has the advantage over [5-3H]uridine because
the hydrogen atom at C6 in uridine exchanges much more slowly
with water than does the C5 hydrogen atom, particularly under
conditions of complete acid hydrolysis of the core nucleotide-
peptide (33).
The protein moiety of the complex must be labelled with
hydrophobic amino acids which are not included in vivo in
nucleic acids and are chemically stable, i.e. Val, Ile, Leu, Phe, Tyr,
and Met. Finally, the protein part of the covalent complex can be
specifically labelled in vitro to high specific activity, for example
with radioactive iodine. The highest incorporation of radioactivity
was obtained after selective acylation of the NH2 group of the
protein (peptide) moiety in the nucleoprotein (34) by unlabeled
-(4-hydroxyphenyl)propionic acid N-hydroxysuccinimide ester
and subsequent iodination of 4-hydroxyphenyl, tyrosine and
histidine residues by the Na125I/chloramine T procedure (35).
In general, the structural analysis consists of three stages:
(i) Isolation of the complex by deproteinization with phenol +
SDS. To prevent loss of the covalent complex during the first
deproteinization step, the concentration of the complex should be
held at a moderate level. At a concentration of nucleic acid
>1 mg/ml, a portion of the complex often partitions to the
interface between the aqueous and organic phases of the
extraction solution. Extraction of the first interface with a small
volume of slightly alkaline buffer containing 0.10.2% SDS
increases recovery of the complex. Generally, after four repeated
extractions, the change in the quantity of protein bound tightly to
nucleic acids is insignificant. Additional precipitation of the
RNAprotein covalent complex with 2 M LiCl is also useful.
(ii) Exhaustive hydrolysis of the complex to produce the smallest
and most homogeneous nucleotide-peptide containing the linkage
unit. Isolation of the nucleotide-peptide from the initial complex
is a key step in structural analysis because it consists of the
covalent linkage unit in which nucleotide and amino acid residues
are bound directly to one another. Two strategies for hydrolysis
of covalent nucleic acidprotein complexes to produce a nucleotide-
peptide core have been used. These include first digesting the
nucleic acid constituent of the complex and then hydrolysing the
protein component. The second strategy merely reverses the
order of hydrolysis. Following either of these two pathways, the
molecular characteristics of the protein or the nucleic acid can
then be determined. Usually proteinase K treatment (2 mg/ml for
several hours) is used to achieve limited digestion of the protein
moiety. Proteolytic digestion of the complex should be carried out
in SDS (0.10.5%) and EDTA (15 mM) to prevent non-specific
hydrolysis of the nucleic acid component. In practice, exhaustive
hydrolysis of RNA in the complex is accomplished using
non-specific endonuclease P1 or S1 and RNase T2 or Pb2.
Treatment of the complex with RNases A and T1 is also productive.
This largely results in a nucleoside diphosphate covalently linked to
protein because the interpolymeric phosphodiester bond is resistant
to RNases. Complete hydrolysis of DNA presents a real challenge
to structural analysis because there is no single nuclease capable
of digesting double-stranded DNA into mononucleotides. DNase
I and exonucleases are mostly used in sequential steps to
hydrolyse DNA. Resistance to 4 M urea, 0.1% SDS and 15%
phenol favours benzonase for identification of proteins tightly bound
to chromosomal DNA (36).
(iii) Identification of nucleotide and amino acid residues directly
bound to one another in the initial complex and determination of
the chemical structure of the covalent bond. As mentioned above,
the core nucleotide-peptide contains a linkage unit where the
phosphate group is shared by nucleotide and amino acid residues.
In the majority of DNA complexes, except for eukaryotic Topo I and
Int family recombinases and vaccinia virus type I topoisomerase
(3740), proteins are bound to the 5-end phosphate group of a
nucleotide. Hydrolysis of the interpolymeric linkage unit with
SVE (for 5 PO derivatives) or spleen phosphodiesterase (for
3 PO derivatives) proves the phosphodiester character of the
bond between nucleic acid and protein in the natural complex. In
addition, it allows identification of the nucleotide residue directly
involved in binding by paper or cellulose thin layer electrophoresis
and/or by ion exchange TLC on DEAE or polyethylenimino
cellulose. In some cases, digestion of the core nucleopeptide with
micrococcal nuclease appeared informative for determination of
its chemical structure. If the linkage unit consists of a nucleoside
diphosphate and a hydroxyamino acid residue, a 3-nucleotide
and O-phosphoamino acid are produced.
To identify amino acids involved in the phosphodiester linkage,
hydrolysis with a mixture of trifluoroacetic and hydrochloric
acids (2:1 v/v, 2030 min, 165
C) is the best choice. The yield of
O-phosphoamino acids is in the range 1116% (25,26,29,30),

good enough for their identification by chromatography or
electrophoresis methods. It is important to keep in mind that
during paper electrophoresis at pH 3.5 pTyr and pU move
together. Thus electrophoretic separation at pH 1.71.9 is
preferred (28).
SOME COMMENTS ON STRUCTURAL ANALYSIS
Until enzymes or chemical agents are identified which specifically
hydrolyse the interpolymeric bond in the covalent complex, the
above-described structural analysis schemes (or variants thereof)
need to be employed. In this context, the hydrolytic properties of
model nucleotide-peptides and activities of the unlinking enzymes,
which specifically hydrolyse interpolymeric phosphodiester bonds,
are considered below.
HYDROLYTIC PROPERTIES OF THE MODEL
NUCLEOTIDE-PEPTIDES
Hydrolysis of phosphoamide and phosphodiester synthetic
derivatives of nucleotides and amino acids (peptides) was
thoroughly examined by Z.A.Shabarova and co-workers in the
1960s1970s (reviewed in 12) and was continued by B.A.Juodkas
laboratory (14,15). A brief summary of selected data important
for structural analysis of covalent nucleic acidprotein complexes
is presented below.
In general, lengthening of the peptide and nucleotide moieties
in model compounds stabilizes phosphodiester derivatives
against hydrolysis by both acid and alkali. Acid hydrolysis of
hydroxyamino acids and nucleotide phosphodiester compounds
occurs under severe conditions (pH <1). In alkali, tyrosine
derivatives are much more stable than serine and threonine
derivatives. When considering the stability or lability to hydrolysis
of the bond formed between nucleic acid and protein, one
important point is the influence of protein (peptide) functional
groups proximal to this bond. In the course of treatment of natural
nucleic acidprotein complexes with alkali, amino groups,
present in positions proximal to the phosphodiester bond formed
both by serine and threonine peptides, can trigger chemical
rearrangements involving the interpolymeric covalent bonds. If
the amino group of serine or threonine peptides is substituted, for
example by an (amino)acyl group, -elimination at the -carbon
atom of the hydroxyamino acid residue in alkaline conditions
takes place at a faster rate than hydrolysis of the interpolymeric
phosphodiester bond. Appearance of a double bond in -aminoa-
crylic acid (in the case of Ser) or in -aminocrotonic acid (in the
case of Thr) can be detected by absorption at 241 nm.
Alternatively, radioactive reagents (Na35SO3 and 14CH3NH2, for
example) added specifically to the double bond may be used. A
free carboxyl group located near the linkage unit suppresses this
-elimination reaction.
Recently, a chemical approach based on -elimination of
VPg (viral protein genome-linked) from nepovirus RNA was
successfully used to identify an N-substituted serine residue
involved in the covalent linkage (41). In this study, the
-elimination reaction could proceed with any serine residue of
VPg except the N-terminal one. If the amino group of Ser or Thr
is free, ON acyl migration from the RNA 5-end nucleotide to
the adjacent NH2 group takes place under alkaline conditions,
i.e. a PO bond is converted to a PN bond.
4793
Nucleic Acids
Nucleic Acids Research,
Research,1994,
1998,Vol.
Vol.22,
26,No.
No.121 4793
UNLINKING ENZYMES
The 5-end uridylic acid of virion (1,2,27) and replicating
(42,43), but not polysomal, picornavirus RNA (44,45) is linked
with the virus polypeptide VPg by a phosphodiester bond via a
unique tyrosine residue. An unlinking enzyme activity,
specifically hydrolysing the interpolymeric phosphodiester bond,
was discovered in animal and plant cells using picornavirus
RNAVPg complexes as substrates (4651). Several possible
functions of the RNA unlinking enzyme which were raised included
triggering of picornaviral RNA translation and regulation of viral
RNA encapsidation or RNA replication (4649). These have not
yet been proved.
An enzyme preparation obtained from HeLa cells was enriched
with a protein with a molecular mass of 29 kDa, however, the low
turnover rate makes its classificaiton as an enzyme problematical
(47).
The substrate specificity of the unlinking enzyme isolated
from ascites Krebs II cells using picorna and comovirus
RNAVPgs, EMC virus RNAVPg derivatives and synthetic
models (34,52) is under examination in our laboratory. We have
developed new efficient substrates (34,51) and methods for
enzyme activity determination (51) to obtain biochemical
enzyme characteristics (34,5153). Briefly, the enzyme is an
acidic protein (pI 4.65.0) (34,53). It exists in cell lysates in a
complex with nucleic acids resistant to dissociation up to 200 mM
KCl (53). Antibodies against the N-terminal dodecapeptide of
EMCV VPg do not inhibit the unlinking enzyme. These results
suggest that the unlinking enzyme recognizes the nucleic acid
constituent of the substrate. The enzyme is active in 0.55 mM
Mg2+. The maximum enzyme activity is in the range 50200 mM
KCl, pH 5.28.0 and 5 mM 2-mercaptoethanol. The enzyme is
stable at pH 4.0, it is inactivated at 55
C and it is inhibited by Ca2+
and Zn2+ at concentrations >1 mM or by 100 mM NH4+ and 20%
acetonitrile.
Further, the unlinking enzyme does not hydrolyse synthetic
phosphodiesters between tyrosine and uridylic acid or tyrosine and
d(TCTCTCCTCTTCCCTCC) (34). Tyr-pT13C2, pT13C2, T13C2,
Cl-phenoxy-pT13C2, CH3O-pT13C2, when added to the substrate in
10-fold molar excess, had little or no effect on the enzyme reaction
(G.Shatskaya, personal communication). Unlinking enzyme
hydrolyses picornavirus and comovirus (here VPg is bound to
RNA via serine residue; 29) RNAVPg substrates differently
(52,54). Keeping in mind these findings, we tentatively named
the unlinking enzyme uridilylpolynucleotide-(5O)-tyrosine
phosphodiesterase (Y-pUpN PDE) (51).
We suggest the following conclusions: (i) it appears that a new
class of cell hydrolases has been found; (ii) since viral substrates
are foreign to non-infected cells, eukaryotic cells harbour
unknown covalent complexes of cellular RNA and proteins
structurally similar to the picornaviral ones.
To determine the role of the unlinking enzyme in picornavirus
infection and to discover its cellular targets, we obtained
antiserum against the EMC virus RNAVPg complex. A fraction
of antibodies which recognized the covalent linkage unit in
either the EMC virus RNAVPg or in synthetic N-Ac-
Tyr(O-pU-2-NH2)OEt and N-Ac-Tyr(O-pT13C2)OEt was found
and partially purified (55). This finding offers promise for using
the picornavirus RNAVPg or synthetic models, mimicking the
covalent linkage unit, as picornaviral vaccines.
4794 Nucleic Acids Research, 1998, Vol. 26, No. 21
Recently, an enzyme which cleaves the covalent bond between
DNA and Topo I in a dead-end complex was found (56). The
DNA unlinking enzyme is presumably involved in repair of DNA
which is impaired by dead-end covalent complex formation
between Topo I and DNA. Hydrolysing the interpolymer
phosphodiester bond, this enzyme releases Topo I from the
covalent complex and generates a free phosphate group, thus
restoring the regular DNA sugarphosphate backbone.
It is vital to note that human Topo I is the sole target of the
campthotecin (CPT) family of anticancer drugs. In the presence
of the drug, the dead-end product described above forms. One can
speculate that the DNATopo I unlinking enzyme might interfere
with CPT therapy, breaking the complex, restoring the function
of DNA and diminishing the therapeutic effect of the drug. Thus,
poisoning of the DNA unlinking enzyme may be desirable for
efficient therapy.
CLASSIFICATION OF NATURAL COVALENT
COMPLEXES OF NUCLEIC ACIDS AND PROTEINS
Currently available data suggest that proteins involved in nucleic
acidprotein complexes can be divided into two classes according
to their properties. Features specific to the two classes of proteins
and their respective complexes are listed below.
Class I
These proteins form covalent complexes with a nucleic acid
through nucleophilic attack by a hydroxyamino acid residue on
an internucleotide phosphate group.
In vitro formation of the covalent complex occurs in a two
component reaction, i.e. nucleic acids and protein.
In the cell, before binding the protein recognizes all aspects of
the nucleic acid structure, i.e. primary, secondary and tertiary.
In the initial stages of the reaction, the protein interacts with the
nucleic acid in a cooperative mode and forms a primary
non-covalent, reversible complex. The cooperative nature of
primary DNA specific complex formation was shown for Topo
I (57), X174 protein A (58) and many other examples of this
class of proteins. The tight (non-covalent) complex formed by
Topo I and DNA (59) or the Col EI relaxation complex (60) can
be dissociated in a high salt Mg2+ solution.
Formation of a tri-substituted phosphate intermediate is possible,
which could induce weakening of the adjacent internucleotide bond.
Decay of this intermediate results in breakage of the internucleotide
bonds on either side of the phosphorous atom. It also depends on
the geometry of the phosphoester bond formed by the hydroxy-
amino acid residue (this was conclusively demonstrated by X-ray
crystallography of the DNAhuman Topo I and DNACre
recombinase intermediate complexes; 23,24). For example, in the
covalent complex, eukaryotic Topo I is bound to a 3-phosphate
group (37), whereas bacterial Topo I is bound to a 5-phosphate
group (61).
Class I proteins are enzymes which combine two activities:
cleavage of an internucleotide bond and restoring its formation.
They belong to the polynucleotide transferase class of enzymes.
The covalent complexes are active for polynucleotide transfer to
an acceptor molecule. Topo I from calf thymus forms a transient
covalent linkage with DNA and reforms the internucleotide bond
in the presence of Mg2+ at 24
C (62). It was reported that
poliovirus VPg (22 amino acid residues), in a covalent complex
with RNA, has RNA cleavagejoining activity (63). However,
our computer analysis (A.Barskii and Yu.Drygin, unpublished
results) with the Genebee program for multiple sequence
alignments revealed no similarity between poliovirus VPg or its
precursor protein (3AB) and the putative active centers of X174
protein A, Topo I or Topo II.
No energy-rich compounds are necessary for either forming the
nucleic acidprotein complex nor restoring the internucleotide
bond nor polynucleotide transfer to an acceptor because the
interpolymeric complex itself has enough free energy to bring
about such reactions. The free energy of the phosphodiester
formed by uridylic acid and tyrosine is 10 kcal/mol (64), an
amount sufficient for internucleotide bond restoration.
It appears that cleavage of the internucleotide bond and
formation of the covalent complex between nucleic acid and
protein take place intramolecularly without the participation of
water (23,24,39).
Class I proteins form so-called relaxed complexes with nucleic
acids. Denaturation of the protein in the primary complex results
in nicking of the nucleotide bond and covalent binding of the protein
to the nascent terminal phosphate group of the nucleic acid.
Members of this class of proteins form covalent complexes in
both the cytoplasm and the nucleus of the cell.
Class II
Members of the second class of proteins do not catalyse covalent
complex formation and its cleavage. Special enzymes do. For
example, the DNA polymerases of bacteriophage 29 (65) and
adenovirus (66) or polio RNA polymerase (67) catalyse covalent
binding of class II proteins to 5-terminal nucleotides of their
respective nucleic acids. In turn, unlinking enzymes catalyse
hydrolysis of interpolymeric phosphodiester bonds between
picornaviral RNAs and VPgs (4652) or between DNA and Topo I
in the dead-end complex (56).
These proteins generate long-lived covalent complexes with
nucleic acids by first interacting with the 5-terminal mono-
nucleoside triphosphate of the nascent polynucleotide in a
multi-component system. Thus, in the complexes formed, the
nucleic acid is always linked to the protein by its 5-end (reviewed
in 1822).
Proteins of this class belong to the so-called terminal proteins
and participate in initiation reactions during nucleic acid replication.
Energy-rich compounds supply the energy for formation of
covalent complexes between Class II proteins and nucleic acids.
The compartments associated with covalent complex formation
for Class II proteins are cellular membranes. Initiation of
replication of nucleic acids in crude cell membrane complexes
has been demonstrated repeatedly (6870).
One can propose that combining the protein properties of a
class might be useful in predicting the formationdisruption
mechanisms of the respective complexes. The classification
above does not pretend to be a final diagnosis; the near future
will show how much these proteins and their respective
complexes are apart.
CELLULAR COVALENT COMPLEXES
Covalent linkage of DNA to cellular proteins involved in DNA
topology changes and DNA recombination has been well
documented (38,71).
Many bacterial plasmids (colicinogenic and sex plasmids and
drug resistance factors) exist as relaxed complexes (72). Treatment

of these complexes with SDS or pronase results in relaxation of the
supercoiled DNA by nicking at the origin of plasmid conjugational
replication and covalent binding of the plasmid-specific protein to
the 5-phosphate group of the nick (73). It should be noted that
relaxed complex formation in bacteria depends on the physiological
status of the cell (74) and is regulated by catabolite repression (75).
It is entirely possible that many eukaryotic circular episomal
DNAs can form relaxed complexes similar to those of bacterial
complexes. This has been shown for the plant Ti plasmid (76). In
addition, one may readily predict that many extrachromosomal
genetic elements [movable genetic elements (77) and cytoplasmic
DNAs (78)] are potential candidates for covalent binding with
proteins.
According to the mechanisms of formationdisjoining of the
covalent complexes, relaxation proteins belong to the class I
proteins discussed above.
Natural tight, presumably covalent, complexes of eukaryotic
chromosomal DNA and cellular proteins have been under intensive
study since the 1980s (8,79,80). There are two questions crucial to
these studies: (i) what are the functions of the proteins involved?;
(ii) where and how are these proteins positioned on DNA?
Although the nucleotide-peptides hydrolysed by phosphodiesterases
have been isolated from animal (8,81) and plant cells (82),
identification of the proteins involved in the linkage remains
obscure. Arrangement of the proteins bound to DNA has mostly
been discussed in the context of the structural organization of
DNA threading in chromatin (82,83). The precise role of
covalently bound proteins in chromosomal organization is not
completely understood. Another issue is what functional component
of chromosomal DNA is a candidate for protein linkage? There
are findings that repeated sequence elements (Alu family) are
tightly linked with proteins (83).
Summing up the data on covalent complexes of chromosomal
DNA with proteins, one can say that proteins with imperfectly
identified molecular characteristics and functions are bound via
phosphodiester bonds with some DNA sequences (8,80,83). It
seems that bulk isolation of the complexes formed by different
proteins makes the structural and functional analyses too
complicated.
To date, a few cellular phosphodiester (presumably) complexes
of RNA and protein have been characterized (8487). The
finding of a p535.8S rRNA covalent complex (84) is a most
exciting observation. The function of this complex is presently
unknown, but the multifunctional properties of this tumour
suppressor protein are well known (88 and references therein).
CONCLUSIONS
The existence of a new family of nucleoproteins in which nucleic
acids are covalently bound to proteins is beyond question. It
should be emphasized that the structure of the complex goes hand
in hand with its function and that successful studies result from
examination of the structure of individual complexes where the
partners were identified. To date, one can predict with certainty
that the number of unknown nucleic acidprotein covalent
complexes is far greater than the number of those known. The aim
of this review has been to provide an overview of the biochemical
approaches to be used to obtain convincing evidence regarding the
covalent nature of nucleoprotein complexes being investigated
now and in the future.
4795
Nucleic Acids
Nucleic Acids Research,
Research,1994,
1998,Vol.
Vol.22,
26,No.
No.121 4795
ACKNOWLEDGEMENTS
I am cordially grateful to Professor A.A.Bogdanov who, back in
the 1960s, involved me in the study of natural covalent complexes
of nucleic acids and proteins. I am indebted to Drs B.Juodka,
Z.Shabarova, Z.Avramova, R.Zanev and D.Werner for helpful
discussions. I deeply appreciate Dr B.L.Semler for critical
reading of the manuscript and English language assistance. I am
thankful to Dr R.Gumport for critical comments. This work was
supported in part by the Russian Foundation for Basic Research
(grants 94-12938 and 97-04-49763) and the G.Soros International
Scientific Foundation (grant M5Z000).
REFERENCES
1 Lee,Y.F., Nomoto,A. and Wimmer,E. (1976) Prog. Nucleic Acid Res.
Mol. Biol., 19, 8996.
2 Flanegan,J.B., Pettersson,R.F., Ambros,V., Hewlett,M.J. and Baltimore,D.
(1977) Proc. Natl Acad. Sci. USA, 74, 961965.
3 Carusi,E.A. (1977) Virology, 76, 380394.
4 Rekosh,D.M., Russel,W.C., Bellet,A.J. and Robinson,A.J. (1977) Cell, 11,
283295.
5 Hermoso,J.M. and Salas,M. (1980) Proc. Natl Acad. Sci. USA, 77,
64256428.
6 Bogdanov,A.A., Antonovich,E.G., Terganova,G.V. and Prokofiev,M.A.
(1962) Biokhimia, 27, 10541060.
7 Drygin,Yu.F. (1971) Thesis, Moscow State University.
8 Juodka,B., Pfutz,M. and Werner,D. (1991) Nucleic Acids Res., 19,
63916398.
9 Bendich,A. and Rosenkranz,H.S. (1963) Prog. Nucleic Acid Res.
Mol. Biol., 1, 219230.
10 Welsh,R.S. and Vyska,K. (1987) Basic Appl. Histochem., 31, 281298.
11 Balis,E.M. and Salser,J.S. (1968) Cancer Res., 28, 595600.
12 Shabarova,Z.A. (1970) Prog. Nucleic Acid Res. Mol. Biol., 10, 145182.
13 Berkowitz,W.F. and Bendich,A. (1965) Biochemistry, 4, 19791983.
14 Juodka,B.A., Kirvielene,V. and Lioroncaite,L. (1979)
J. Carbohydrates Nucleosides Nucleotides, 6, 333357.
15 Juodka,B.A., Kirvielene,V. and Povilionis,P. (1982)
Nucleosides Nucleotides, 1, 111125.
16 Wang,J.C. (1971) J. Mol. Biol., 55, 523533.
17 Eisenberg,S., Griffith,J. and Kornberg,A. (1977) Proc. Natl Acad. Sci.
USA, 74, 31983202.
18 Juodka,B.A. (1980) Bioorg. Khim., 6, 14451465.
19 Salas,M. (1980) Trends Biochem. Sci., 5, 191193.
20 Drygin,Yu.F. (1982) Covalently Bound Proteins and Nucleic Acids
A New Class of Natural Compounds (in Russian), Annual Review of
Science and Technology, Molecular Biology Series. VINITI, Moscow,
Russia, pp. 139189.
21 Wimmer,E. (1982) Cell, 28, 199201.
22 Vartapetian,A.B. and Bogdanov,A.A. (1987) Prog. Nucleic Acid Res.
Mol. Biol., 34, 209251.
23 Guo,F., Gopaul,D.N. and Van Duyne,G.D. (1997) Nature, 389, 4046.
24 Redinbo,M.R., Stewart,L., Kuhn,P., Champoux,J.J. and Ho,W.G.H. (1998)
Science, 279, 15041513.
25 Zolotukhin,A.S., Drygin,Yu.F. and Bogdanov,A.A. (1984) Biochem. Int., 9,
799806.
26 Zolotukhin,A.S., Drygin,Yu.F. and Bogdanov,A.A. (1985) Proc. Acad. Sci.
USSR, 283, 213215.
27 Drygin,Yu.F., Vartapetian,A.B. and Chumakov,K.M. (1979) Mol. Biol., 13,
777787.
28 Vartapetian,A.B., Drygin,Yu.F., Chumakov,K.M. and Bogdanov,A.A.
(1980) Nucleic Acids Res., 8, 37293742.
29 Drygin,Yu.F., Sapotsky,M.V. and Bogdanov,A.A. (1987) FEBS Lett., 215,
247251.
30 Drygin,Yu.F., Zuklys,K.L., Terskich,A.V. and Bogdanov,A.A. (1988) Eur.
J. Biochem., 175, 5763.
31 Sburlati,A.R., Monrow,R.E. and Berger,S.L. (1991) Proc. Natl Acad. Sci.
USA, 88, 253257.
32 Drygin,Yu.F., Kozlov,I.A. and Bogdanov,A.A (1970) Biokhimia, 35,
10141022.
4796 Nucleic Acids Research, 1998, Vol. 26, No. 21
33 Kochetkov,N.K., Budovsky,E.I, Sverdlov,E.D., Simukova,A.,
Turchinsky,M.F. and Shibaev,V.N. (1970) Organic Chemistry
of Nucleic Acids. Khimia, Moscow, Russia, pp. 146215.
34 Shabanov,A.A., Kalyuzny,A.A., Gottikh,M.B. and Drygin,Yu.F. (1996)
Biokhimia, 61, 11061118.
35 Bolton,A.E. and Hunter,W.M. (1973) Biochem. J., 133, 529539.
36 Juodka,B., Spiess,E., Angiolillo,A., Joswig,G., Rothbarth,K. and
Werner,D. (1995) Nucleic Acids Res., 23, 13591366.
37 Champoux,J. (1977) Proc. Natl Acad. Sci. USA, 74, 38003804.
38 Nash,H. (1977) Curr. Topics Microbiol. Immunol., 78, 171199.
39 Petersen,B.O. and Shuman,S. (1997) Nucleic Acids Res., 25, 20912097.
40 Esposito,D. and Scocca,J.J. (1997) Nucleic Acids Res., 25, 36053614.
41 Zalloua,P.A., Buzayan,J.M. and Bruening,G. (1996) Virology, 219, 18.
42 Takegami,T., Semler,B.L., Anderson,C.W. and Wimmer,E. (1983)
Virology, 128, 3347.
43 Vartapetian,A.B., Koonin,E.V., Agol,V.I. and Bogdanov,A.A. (1984)
EMBO J., 3, 25832588.
44 Hewlett,M.J., Rose,J.K. and Baltimore,D. (1976) Proc. Natl Acad. Sci.
USA, 73, 327330.
45 Nomoto,A., Lee Y.F. and Wimmer,E. (1976) Proc. Natl Acad. Sci. USA,
73, 375380.
46 Ambros,V., Petterson,R.F. and Baltimore,D. (1978) Cell, 15, 14391446.
47 Ambros,V. and Baltimore,D. (1980) J. Biol. Chem., 255, 67396744.
48 Dorner,A.J., Rothberg,P.E. and Wimmer,E. (1981) FEBS Lett., 132, 219223.
49 Sangar,D.V., Bryant,J., Harris,T.J.R., Brown,F. and Rowlands,D.J. (1981)
J. Virol., 39, 6774.
50 Drygin,Yu.F. and Syuanova,E.Yu. (1986) Biokhimia, 51, 249259.
51 Drygin,Yu.F., Shabanov,A.A. and Bogdanov,A.A. (1988) Biokhimia, 53,
13711379.
52 Drygin,Yu., Shabanov,A. and Bogdanov,A. (1988) FEBS Lett., 239,
343346.
53 Viryasov,M.B., Yusupova,R.A., Shatskaya,G.S. and Drygin,Yu.F. (1997)
Chromatographia, 45, 133137.
54 De Varennes,A., Lomonosoff,G.P., Shanks,M. and Maule,A.J. (1986)
J. Virol., 27, 23472351.
55 Bordunova,O., Turina,O., Nadezhdina,E., Shatskaya,G., Veiko,V. and
Drygin,Yu. (1998) FEBS Lett., 422, 5760.
56 Shu-wei Yang, Burgin,A.B.,Jr, Huizenga,B.N., Robertson,C.A., Yao,K.C.
and Nash,H.A. (1996) Proc. Natl Acad. Sci. USA, 93, 1153411539.
57 Liu,L.F. and Wang,J.C. (1979) J. Biol. Chem., 254, 1108211088.
58 Ikeda,J.-E., Yudelevich,A., Shimamoto,N. and Hurwitz,J. (1979)
J. Biol. Chem., 254, 94169430.
59 Depew,R.E., Liu,L.F. and Wang,J.C. (1978) J. Biol. Chem., 253, 511518.
60 Blair,D.G. and Helinski,D. (1975) J. Biol. Chem., 250, 87858789.
61 Tse,Y.-C., Kirkegaard,K. and Wang,J.C. (1980) J. Biol. Chem., 255,
55605565.
62 Been,M.D. and Champoux,J.J. (1981) Proc. Natl Acad. Sci. USA, 78,
28832887.
63 Tobin,G.J., Young,D.C. and Flanegan,J.B. (1989) Cell, 59, 511519.
64 Holzer,H. and Wohlhueter,R. (1972) In Weber,G. (ed.), Advances in
Enzyme Regulation. Pergamon Press, Oxford, UK, Vol. 10, pp. 121132.
65 Blanco,L. and Salas,M. (1996) J. Biol. Chem., 271, 85098512.
66 Kenny,M.K., Balogh,L.A. and Hurwitz,J. (1988) J. Biol. Chem., 263,
98019808.
67 Paul,A., van Boom,J.H., Filippov,D. and Wimmer,E. (1998) Nature, 393,
280284.
68 Takegami,T., Semler,B.L., Anderson,C.W. and Wimmer,E. (1983)
Virology, 128, 3347.
69 Vartapetian,A.B., Koonin,E.V., Agol,V.I. and Bogdanov,A.A. (1984)
EMBO J., 3, 25832588.
70 Takeda,N., Kuhn,R.J., Young,C.-F., Takegami,T. and Wimmer,E. (1986)
J. Virol., 60, 4353.
71 Wang,J.C. (1996) Annu. Rev. Biochem., 65, 635692.
72 Clewell,D.B. and Helinski,D. (1969) Proc. Natl Acad. Sci. USA, 66,
11591166.
73 Bastia,D. (1978) J. Mol. Biol., 124, 601639.
74 Womble,D.D. and Rownd,R.H. (1977) J. Bacteriol., 131, 145152.
75 Katz,L., Kingsbury,D.T. and Helinski,D. (1973) J. Bacteriol., 114, 577591.
76 Ward,E.R. and Barnes,W.M. (1988) Science, 242, 927930.
77 Sakaguchi,K. (1990) Microbiol. Rev., 54, 6674.
78 Abken,H., Hegger,R., Butzler and Wilecke,K. (1993) Proc. Natl Acad. Sci.
USA, 90, 65186522.
79 Neuer,B., Plagens,B. and Werner,D. (1983) J. Mol. Biol., 164, 213235.
80 Avramova,Z., Ivanchenko,M. and Tsanev,R. (1988) Plant Mol. Biol., 11,
401408.
81 Neuer-Nitsche,B. and Werner,D. (1987) Biochem. Biophys. Res. Commun.,
147, 335339.
82 Tsanev,R. and Avramova,Z. (1994) Chromosoma, 103, 293301.
83 Werner,D. and Neuer-Nitsche,B. (1989) Nucleic Acids Res., 17, 60056015.
84 Fontoura,B.M., Sorokina,E.A., David,E. and Carrol,R.B. (1992)
Mol. Cell. Biol., 12, 51455151.
85 Caizzi,R. and Ritossa,F. (1983) Biochem. Genet., 21, 267285.
86 Ahn,Y.S., Choi,Y.C., Goldknopf,I.L. and Bush,H. (1985) Biochemistry, 24,
72967304.
87 Thomas,L., Pfeifle,J. and Anderer,F.A (1987) Biochim. Biophys. Acta, 909,
173178.
88 Ko,L.J. and Prives,C. (1996) Genes Dev., 10, 10541072.