WO1989002475A1 - Polyaldehydic polynucleotides in use as probes, their preparation and use - Google Patents

Abstract

The invention provides a polynucleotide probe having a base sequence hybridizable to a preselected polynucleotide and also having a multiplicity of reporter groups each linked thereto via an aldehyde group which has been provided by oxidation of a non-backbone sugar residue. The use of aldehyde groups enables post-hybridization labelling to be achieved easily with great benefits in analysis/diagnosis.

Description

POLYALDEHYDIC POLYNUCLEOTIDES IN USE AS PROBES. THEIR PREPARATION AND USE

This invention relates to polyaldehydic polynucleotides in use as probes, their preparation and use. The invention is particularly concerned with, but not limited to, polynucleotides having reporter groups attached thereto. In the following, literature references are indicated by numbers in parenthesis.

Modified RNA and DNA molecules may be useful in a number of practical applications. Many of the currently envisaged applications involve hjybridization of these modified nucleic acid molecules (hereafter called "probes") to complementary DNA and RNA sequences (hereafter called "targets"). This class of potential applications includes DNA-based diagnostics, use as anti-messages to modulate levels of gene expression (1), and as a component of novel methods for cleaving DNA or RNA chains at any specific site (2). Usually, the probes are modified prior to hybridization with the target. For application to DNA-based diagnostics, entities added to the DNA probe have been called reporter groups. They function, inter alia, as "labels" to permit the probes to be traced. For most practical purposes, non-radioactive reporter groups are preferred (3). Although the term "reporter" does not adequately describe all the functions envisaged for the entities added to the nucleic acid molecules, this term is used loosely herein to refer to any entity artificially attached, covalently or not, to the probe molecules. Usually, the probe molecules are labelled with the reporter groups prior to hybridization with the target sequences. This is done because it is usually difficult selectively to label only the probe component of the probe-target hybrid molecules after hybridization (some exceptions are discussed below).

It is generally recognized that reporter groups attached to nucleic acid probes prior to hybridization with the target sequences should fulfill the following five criteria:-

(1) The reporter group should not interfere with the hybridization process to any serious extent.

(2) The reporter group should not decrease the thermal stability of the probe-target hybrid molecules.

(3) The reporter group must itself be stable during the hybridization process.

(4) The bond between the reporter group and the probe must resist the hybridization procedure.

(5) The reporter group must not induce non-specific binding of the probe to the target sequences.

An important practical consequence of these requirements is that methods employing probes labelled with non-radioactive reporter groups are usually not as sensitive as methods employing radioactivity labelled probes. That is, target sequences are not as rapidly or as specifically detected as with, say, nick translated, radiolabelled probes (4).

Ideally, the art requires a method of producing non-radioactivity labelled probes that are even more sensitive than the radiolabelled probes. Consequently, the attainment of ever higher levels of sensitivity is a major and continuing effort in the field of DNA-based diagnostics. However, with currently available technologies, there appears to be a limit to the number of reporter groups that can be added to the probes prior to hybridization without interferring with one or more of the five requirements mentioned above. These limitations thus create a general problem in applications involving nucleic acid hybridizations that can be stated as follows: how to increase the signal without increasing noise.

Some of the limitations of the currently available methodologies can be illustrated with biotinylated probes. In these probes the biotin is present prior to hybridization. However, in practice one never achieves the theoretically maximum sensitivity, which would be to attach biotin to all of the biotin-accepting bases in the probe molecule. That is, in currently available methods, the biotin is not attached to all of the biotin-accepting bases (37, 38). Thus, biotin as a reporter group illustrates a general problem associated with currently available probes: how can one attach reporter groups to all of the biotin (or reporter)-accepting bases in a probe molecule, particularly post hybridization. In contrast to previously described probes, the present invention solves the problem of how to attach reporter groups to all of the reporter-accepting bases and thus solves the general problem of how to achieve the theoretical maximum sensitivity. Consequently, with the present invention, one can attach biotin to all of the biotin-accepting bases in a way that avoids the problems encountered when biotin is added to all of the biotin-accepting bases prior to hybridization.

The principles and methods disclosed here offer a general solution to this particular problem. These principles and methods provide the basis for an approach in which virtually unlimited numbers of reporter molecules can be covalently attached specifically to the probe molecules after (instead of before) they have hybridized to the target sequences. Furthermore, and as will become clear, these principles, methods and the products produced thereby have applications in other fields as well.

European Patent Specification No. 0133473 is concerned with in vivo labelling of polynucleotide probes and in Example II thereof describes, in isolation, the biotinylation of an unidentified "T4 DNA". Although the method described uses periodate oxidation, the method is not correlated in any way with hybridization work nor with a particular probe or probes. Neither is described. Hybridization data is critical to use of any material in or as a polynucleotide probe, and there is no such data present. It is not even possible to be certain of the nature of the "T4 DNA" used and there is no disclosure whatsoever of its use.

Thus, it is not possible to deduce from this disclosure whether or not the polyaldehydic DNA of this invention would itself be a useful probe, nor is it possible to deduce the unique analytical/ diagnostic methods of the invention referred to hereinafter. The principles and methods disclosed herein can produce single or double stranded nucleic acid molecules containing large numbers of chemically reactive aldehyde groups. These aldehydes are produced from a suger (e.g. glucose) component of glucosylated polynucleotides, e.g. DNA. In the art, with the exception of the European patent specification referred to above, no previously described nucleic acid molecules contain more than a single pair of chemically reactive aldehyde groups and these, in contrast to the present invention, are derived from a sugar component of the nucleic acid and not from non-backbone residues. These are RNA molecules in which the ribose moiety at the 3' end of the polynucleotide backbone has been oxidized (5). They will be discussed further below.

Numerous kinds of compounds can be covalently attached by chemical methods to the aldehydes of the polyaldehydic products produced according to this invention.

The specificity of the chemical reactions disclosed here suggests that the practical problems encountered with DNA probes that have been discussed above can be solved in a novel and general manner. This is evident if one considers using probe molecules composed of glucosylated DNA. The glucosylated DNA can be efficiently converted to polyaldehydic DNA by the principles and methods disclosed here. Non-glucosylated DNA is not converted to polyaldehydic DNA by these methods. Thus, with glucosylated DNA probes, one can first hybridize the (glucosylated) probe with the (non- glucosylated) target sequences and then convert only the probe strand of the probe-target hybrid molecules into a polyaldehydic DNA chain.

Far more important, the invention enables conversion to aldehydic form before hybridization (followed by labelling with reporter groups). The advantages of this technique are considerable. First, the end user of the probe does not have to employ an oxidation step; a pre-oxidized probe can be used.

Secondly, the presence of possibly harmful oxidizing agents in any of the solutions containing probe-target hybrid molecules is avoided thus preventing any interference with hybridization or adverse effects on any substrate used subsequent to hybridization. Thirdly, for the first time in an unexpected fashion RNA sequences may be detected by post-hybridization labelling techniques that employ covalent bonds. Not only is labelling prior to hybridization avoided, but chemical attack, of the RNA is also avoided since a pre-oxidized probe is used. This is an important consideration because oxidation and labelling after hybridization with a population of RNA. molecules would usually result in the oxidation of the ribose moiety at the 3' ends of RNA chains and the production of a single pair of aldehydes. Since RNA molecules that are not "targets", and which are in vast excess over the true "targets", would thus also become labelled with the aldehyde reactive reporter groups, there would be an increased level of "noise" resulting from. this non-specific labelling.

When the target molecules are DNA, one can use either approach; that is, hybridize first and then oxidize or first oxidize the probe molecules and then hybridize. In either way one can attached the reporter or other molecules after hybridization specifically only to the probe strand of the probe-target hybrid molecules. Since the reporter molecule can be attached to the probe at a much lower temperature than was used for the hybridization, reporter-probe-target complexes can be produced which will be stable at, say, room temperature, but which would have been very unstable had they been formed at thβά elevated temperatures required for hybrid formation. One consequence of this approach is that there may be no limit on the number of reporter molecules that can be attached to each probe molecule. Thus, a novel and general solution to the problem of the limited sensitivity of non-radioactive probes is provided in an unsuspected way which is astonishingly simple.

This invention unexpectedly provides new opportunities for increasing the sensitivity of detection employing previously used DNA detection systems, new methods, compounds and chemistry for linking reporter groups to DNA probes and new opportunities for developing entirely novel detection systems for diagnostics. These possibilities can be seen from the following illustrative examples. Detection systems, or components of detection systems exploited previously only in protein-based diagnostics can be used. For example, one of the antigenic components of such systems, such as insulin, could be directly attached to the polyaldehydic probes after hybridization. The ability to produce pairs of aldehydes very close to one another on a DNA chain should make it possible to significantly increase the signal in systems involving two components that must physically be closely associated, such as energy transfer systems (34).

A related idea should permit a significant enhancement, not of signal strength itself, but of the signal-to-noise ratio. A characteristic feature of probes labelled according to this invention is that, distributed along the chain according to the nucleotide sequence, some labels will when seen in three dimensions, be in pairs of clusters.

One could also construct a "tree" of reporter groups on each aldehyde by employing polymers previously unused in DNA-based diagnostics, containing, for example, but not limited to, amino acids (e.g., derivatives of polylysine or polyglutamic acid) or sugar moieties (e.g., polysaccharides or even another polyaldehydic DNA molecule) or amino-alcohols. These and other kinds of polymers could be attached to the polyaldehydic probe DNA after hybridization by attaching a reactive o-alkyl-hydroxylamine to one end of the polymer. Numerous reporter groups could be added to each polymer either before or after the polymer had been attached to the probe. In these particular examples, the reporter groups could be attached either directly to the amine groups on the polylysine or to the aldehydes created after oxidation of the polysaccharide or poly amino-alcohol chains.

O-alkyl-hydroxylamine reagents are particularly attractive because they are highly specific for aldehydes, because they form adducts rapidly and because, unlike reagents containing aliphatic or aromatic amines, they form stable adducts with polyaldehydic DNA without the use of additional reagents, e.g., cyanoborohydride, to reduce the Schiff-base intermediate between aldehydes and amines.

Polyaldehydic probes also provide the opportunity to create entirely new detection and amplification systems. For example, one could attach compounds which can act as nucleation centers for the stable polymerization of protein or other kinds of monomers into large structures. One example of such a compound is Phalloidin and chemically related compounds, which can nucleate the polymerization of actin monomors into long filaments (F-actin) (35). This would create a bundle of actin filaments around each probe-target hybrid molecule. Since actin bundles can be seen in the light microscope, one might be able, using either native or fluorescent labelled actin monomers, to quantitatively determine the absolute number of target molecules in the sample being analyzed. In addition, enzymes that have not been successfully linked to DNA probes via avidin-biotin systems, such as luciferases, could be linked to DNA through the aldehydes on polyaldehydic probes. Thus, polyaldehydic probes can provide a number of surprising advantages that were previously unrealized with any kind of DNA probe.

Another surprising advantage of the methods of this invention is that the aldehydes can be selectively produced, favoring those probe molecules that have hybridized with the target sequences. Thus, this invention can both simplify the procedure used in the diagnostic process and provide a new and novel solution to the problem of how to reduce the noise (the non-specific signal) coming from the excess, unhybridized probe molecules still present after the hybridization step. These advantages can be appreciated when one considers that, in the diagnostic test, the concentration of the probe molecules is generally in excess over the concentration of the target molecules. Consequently, in order to be able to detect the probe-target hybrids, it is usually necessary to perform a purification step in which the unhybridized probe molecules are separated from the probe-target hybrids. However, if the reporter groups could be attached preferentially to those probe molecules that were in a hybrid with the target, this separation step could be eliminated entirely from the diagnostic procedure, or as a minimum, one would have a reduced level of "noise" after the separation step. This can be done according to the present invention by hybridizing with probe molecules that contain the glucose-accepting base, 5-hydroxymethylcytosine (HMC), but which are not linked to either glucose or to the aldehyde pairs produced by oxidation of the glucose. The glucosyl transferase enzymes, which transfer glucose from UDP-glucose to the glucose-accepting bases, preferentially add glucose to HMC bases that are in double stranded DNA over HMC bases that are in single stranded DNA (17, 18). Thus, one can hybridize first, then add glucose preferentially to the double stranded probe-target hybrid molecules and then oxidize the non-backbone sugar moieties. In this way, the aldehydes will be produced preferentially on those probe molecules that have hybridized with the targe sequences. Since the aldehyde-reactive reporter groups will not efficiently attach to the unhybridized probe molecules, one could just ignore the unhybridized probe molecules and not separate them from the probe-target hybrids. Alternatively, one could still perform the separation step and have a reduced level of "noise". Thus, this method can simplify the procedure, reduce the noise level and solve the problem of how to remove the unhybridized probe molecules. When employing this method in some diagnostic situations, it may be necessary to separate the complementary probe strands from one another and to then hybridize with only one of the strands, in order to avoid having complementary probe strands hybridize with one another and thus eventually become attached to the reporter groups.

The ability to produce polyaldehydic DNA also offers an entirely different and previously unexplored chemical basis for attaching reporter or other molecules to DNA probes prior to hybridization with target sequences. Since the physical and chemical properties of polyaldehydic DNA are different from any previously employed DNA probes, molecules previously considered unsuitable as reporter molecules with other types of probes might well be usable with polyaldehydic probes, e.g. molecules which are incompatible with the use of DNA polymerases can be employed and molecules which are incompatible with the conventional use of blocking groups in nucleotide synthesis. In general, reporter groups can now be used which cannot be used prehybridization. The different chemical options provided by the presence of aldehyde groups opens up new possibilities.

Probes made of glucosylated DNA (which offers the possibility of specifically labelling the probe after hybridization) would differ in two important respects from the superficially similar situation encountered with RNA-DNA hybrids. RNA-DNA hybrids are immunologically distinct from either RNA-RNA or DNA-DNA duplexes and thus can also be specifically labelled after the hybridization step (7). With the methods of the present invention, the reporter molecules would not be added to the hybrid per se, but rather only to the aldehyde groups present uniquely on the probe strand of the probe-target hybrids and they can be added chemically, not immunologically . The present invention applied to DNA-diagnostics is also distinct from situations where bases on the probe molecules can be chemically modified prior to hybridization to make them antigenic (8,9). In these situations, entities can be attached directly to the probe strand of the probe-target hybrids, but only by immunological methods.

Of course, the potential applications of polyaldehydic polynucleotides prepared according to the principles and methods of this invention are not limited to DNA-based diagnostics and other situations involving nucleic acid hybridizations. In these applications, the polyaldehydic polynucleotide is potentially useful because it can perform two functions: it can be part of a duplex molecule with complementary nucleic acid sequences and it can serve as a physical structural support for the attachment of reporter molecules. In some embodiments of this invention, the important quality of the polyaldehydic products may primarily be their ability to function ass as physical structure or support for the attachment of various entities or for linking various entities. For example, polyaldehydic DNA can be used as a support for immobilizing enzymes or as a protein cross linking agent. Wool cross linked with bifunctional aldehyde molecules is reported to have altered properties (10). Wool fibres cross linked by polyaldehydic DNA may have other novel and useful properties. Polyaldehydic DNA can also serve as a polyvalent aldehydic linker for the attachment of numerous toxin or other useful entities to monoclonal antibodies (6), opening up the prospect of many new drugs and therapeutics by employment of techniques in principle well known in the field of immunotoxins or other "target specific" drugs.

Thus, the invention provides a polynucleotide probe having a base sequence hybridizable to a preselected polynucleotide and also having a multiplicity of reporter groups each linked thereto via an aldehyde group which has been provided by oxidation of a non-backbone sugar residue.

The invention also includes a duplex comprising a polynucleotide hybridized to a polynucleotide probe designed to detect a chosen base sequence or polynucleotide, the probe carrying a multiplicity of aldehyde groups or a multiplicity of reporter groups each linked to the probe by an aldehyde group, and, in either case, the aldehyde groups each being provided by oxidation of a non-backbone sugar residue.

A further aspect of the invention is a duplex comprising a polynucleotide hybridized to a polynucleotide probe designed to detect a chosen base sequence or polynucleotide, the probe carrying intact non-backbone sugar residues so as to provide means for identifying the duplex by oxidizing said sugar residues to provide aldehyde groups to which reporter groups may be attached.

Further provided by the invention is a process for preparing a polynucleotide probe which comprises providing a polynucleotide having a base sequence hybridizable to a preselected polynucleotide and also in which there are non-backbone sugar residues and oxidising the said residues to provide aldehyde groups.

In yet another aspect, the invention includes a non-T4 derived polyaldehydic polynucleotide having aldehyde groups provided by at least one oxidized non-backbone sugar residue. Such polynucleotides (or indeed those of T4 origin) can be used, e.g. as a support for an immobilized enzyme, as a protein cross-linking agent, or as a aldehydic linker, e.g. wherein the polynucleotide is employed to link a toxin to a monoclonal antibody or to link a molecule having a biological effect on target cells to a monoclonal antibody for those cells.

The invention provides a method of diagnosis not being one practised on the human or animal body and in which a probe as defined above is employed to detect by hybridization a polynucleotide characteristic of a disease or disorder or stage thereof.

The invention also includes the following methods:

(a) A method of diagnosis in which a polynucleotide having aldehyde groups provided by at least one oxidized non-backbone sugar residue and which is hybridizable to a polynucleotide characteristic of a disease or disorder or stage thereof is subjected to hybridization with a sample suspected to contain said characteristic polynucleotide, and any hybrids thus-formed are labelled by attachment of reporter groups to said aldehyde groups and thereafter detected using said reporter groups.

(b) A method of diagnosis in which a polynucleotide probe having non-backbone sugar residues and being hybridizable to a polynucleotide sequence characteristic of a disease or disorder or stage thereof is subjected to hybridization with a sample suspected to contain said characteristic polynucleotide, after hybridization aldehyde groups are provided on the probe by oxidizing said non-backbone sugar residues, and any hybrids which have been formed are labelled by attachment of reporter groups to said aldehyde groups and thereafter detected using said reporter groups.

(c) A method of diagnosis in which a polynucleotide probe which is hybridizable to a polynucleotide characteristic of a disease or disorder or stage thereof is subjected to hybridization with a sample suspected to contain said characteristic polynucleotide, said probe is thereafter provided with at least one non-backbone sugar residue and said residue(s) oxidized to provide aldehyde groups, and any hybrids which have been formed are detected by attachment of reporter groups to said aldehyde groups

(d) A method of analysis in which a polynucleotide having aldehyde groups provided by at least one oxidized non-backbone sugar residue and which is hybridizable to a polynucleotide which it is desired to detect is subjected to hybridization with a sample suspected to contain said polynucleotide to be detected, and any hybrids thus-formed are labelled by attachment of reporter groups to said aldehyde groups and thereafter detected using said reporter groups.

(e) A method of analysis in which polynucleotide probe having non-backbone sugar residues and being hybridizable to a polynucleotide which it is desired to detect is subjected to hybridization with a sample suspected to contain said polynucleotide to be detected, after hybridization aldehyde groups are provided on the probe by oxidizing said non-backbone sugar residues, and any hybrids which have been formed are labelled by attachment of reporter groups to said aldehyde groups and thereafter detected using said reporter groups.

(f) A method of analysis in which a polynucleotide probe which is hybridizable to a polynucleotide which it is desired to detect is subjected to hybridization with a sample suspected to contain said polynucleotide to be detected, said probe is thereafter provided with at least one non-backbone sugar residue and said residues oxidized to provide aldehyde groups, and any hybrids which have been formed are detected by attachment of reporter groups to said aldehyde groups.

The nucleic acid used for specifically demonstrating the principles disclosed herein is biologically produced glucosylated DNA. However, the scope of this invention is not limited to this kind of nucleic acid molecule. Suitable DNA (or RNA) molecules can be produced entirely or partially in vitro. Since these synthetically produced nucleic acids can be treated according to the principles disclosed here, they are included in the scope of this invention. We will briefly describe some approaches that can be employed to produce suitable DNA or RNA molecules in vitro.

In vivo, glucosylated DNA of the T-even family of bacteriophages is produced as the end result of a complicated enzymatic process that can be summarized as involving two distinct steps: (1) incorporation of a glucose-accepting nucleotide into a polynucleotide chain followed by (2) the covalent addition of one or two glucose molecules to the glucose-accepting nucleotide. The critical consideration for this invention is that either of these steps can be performed in vitro. The glucose accepting nucleotide contains the unusual base 5'-hydroxymethylcytosine (HMC). In the DNA of the T-even bacteriophages (exemplified by bacteriophage T4), HMC replaces all of the cytosines that are normally present in most DNA species. In vivo, glucose can be attached to the 5' -hydroxymethyl group of these HMC bases through the action of enzymes that are called glucosyl transferases.

We will first discuss the in vitro production of polynucleotides containing the potential glucoseaccepting base, HMC, then some properties of the glucosyl transferase enzymes and how they can be employed, and finally how sugars other than glucose can be added to polynucleotide chains containing glucose-accepting bases.

The in vitro enzymatic polymerization of hydroxymethylcytosine (HMC) into polydeoxyribo- nucleotide chains was first reported in 1961 (11). Thus, one can produce, in vitro, using known techniques, DNA chains that contain a potential glucose-accepting nucleotide but which do not actually have any glucose attached to the DNA. DNA containing HMC, but having little or no glucose attached, has also been produced biologically (12). DNA containing HMC bases can also be produced chemically (13).

In some embodiments of this invention, it might be advantageous to have both cytosine and hydroxymethylcytosine in the same polynucleotide chain. If it is important precisely to control the position(s) of the glucose-accepting HMC bases along the nucleic acid chain, synthetically produced polynucleotide chains might be preferable to those produced biologically. That is, only with in vitro methods would it always be possible precisely to control where the glucose accepting and non-accepting cytosines would be located. Glucose can then be added, in vitro, to these synthetically produced DNA molecules. Thus, this invention relates to polynucleotides containing one or more, oxidizable, non-backbone, sugar moieties.

In vivo, glucose is attached to DNA chains containing the unusual base, hydroxymethylcytosine (HMC). However, in some embodiments of this invention it might be preferable to employ RNA chains instead of DNA chains. Such a situation could arise, for example, in applications involving nucleic acid hybridizations. Thus, under certain conditions, RNA-DNA hybrids are formed preferentially over DNA-DNA hybrids (14). That is, under these conditions, RNA molecules can efficiently hybridize to complementary DNA sequences while the hybridization of DNA molecules to complementary DNA sequences is minimized. Consequently, it can sometimes be useful to produce RNA chains containing the glucose accepting base, 5'-hydroxymethylcytosine (HMC). RNA chains containing modified nucleotides can be produced enzymatically and chemically (13). They might also be produced in vitro with some RNA polymeases or synthetases. That is, some RNA polymerizing enzyme might incorporate ribonucleotides bearing the unusual base HMC into polynucleotide chains. Glucose might then be added in vitro to these synthetically produced RNA chains by a glucosyl transferase enzyme. Although RNA chains containing the glucose-accepting base, HMC, have not been produced, one can reasonably anticipate that a glucosyl transferase enzyme, or a mutant varient of such an enzyme, will be able to transfer glucose, under standard or altered reaction conditions, to 5'hydroxymethyl-cytosine containing RNA chains that are in a single stranded state or in a hybrid structure with either another RNA chain or in a hybrid with a DNA chain. That is, the activity of these enzymes, which requires a modified base, may be relatively indifferent to whether the backbone of the polynucleotide chains are composed of ribose or deoxyribose.

The glucose accepting base, HMC is a pyrimidine carrying a hydroxymethyl group at the 5' position of the pyrimidine ring. Recognition of this 5' hydroxy methyl group must therefore be important for the activity of the glucosyl transferase enzymes. In a polynucleotide chain, the presence of a 5' hydroxymethyl group on a pyrimidine ring may be the only or the major structural requirement for the activity of one or more of these transferase enzymes. Thus, other 5' hydroxymethyl-pyrimidine bases (for example, 5' hydroxymethyluracil, a natural constituent of the DNA of a number of Bacillis subtilis phages (15)), whether in DNA or RNA chains, might also serve as glucose acceptors for these transferase enzymes.

It should be stated here that the glucose of glucosylated DNA does not interfere with nucleic acid hybridizations. That is, glucosylated DNA is frequently used in nucleic acid hybridizations (16) and has never been reported to interfere with the hybridization process. It should also be pointed out that glucose could be added enzymatically to the (5'-hydroxymethylcytosine-containing) probe molecules after they have formed hybrids with the target nucleic acid sequences. That is, the glucosyl transferase enzymes could be employed after hybridization of probe and target sequences as well as during preparation of the probe molecules. This order of operations might be preferable for situations where the probe molecules are produced chemically. This is because chemically produced polynucleotides are single stranded and because most of the known glucosyl transferase enzymes are significantly less active on single stranded DNA than on double stranded DNA (17,18).

Three types of glucosyl transferase enzymes were identified and partially characterised in 1961 and 1962 (16,17,18). Recently, most of the genes encoding these enzymes have been cloned (19,20), and some of them have been expressed at high levels in E. coli (19). These enzymes are encoded in the bacteriophages genomes. With two types of these enzymes a monosaccharide is produced: single glucose molecules are covalently attached to the 5' position of the grlucose-accepting base, hydroxymethylcytosine inaan aipha- or in a beta-O-glycosidic linkage. With the third type a disaccharide is produced: a single glucose as attached in a beta-linkage on the number six carbon of a glucose that is already attached to the DNA. This disaccharide is gentiobiose (21). All of these enzymes are active in. vitro and have been used to produce glucosylated DNA whose glucosylation pattern is different from the in vitro pattern on T-even bacteriophage DNA's (11,18). The critical consideration for the present invention is that these enzymes can glucosylate, in vitro, any DNA sequence that contains the glucose-accepting base, HMC. That is, in vitro, these enzymes will glucosylate DNA other than T-even bacteriophage DNA if this DNA contains the potential glucose accepting nucleotide. Thus, glucosylated DNA of any desired sequence can be produced: in vitro.

Although these glucosyl transferase enzymes are welll known, they are not frequently used in the scientific community. They have been only partially characterized (20) and they have not been the subject of any studies designed to alter or modify the specificity of their enzymatic activity. In particular, it is not known if any of these enzymes or a mutationally altered form of any of these enzymes, might, under some conditions add a sugar other than glucose to DNA containing a 5' hydroxymethyl-cytosine base. That is, it is not presently known if the sugar donor molecule can be something other than UDP-glucose (11). Thus, it is reasonable to anticipate that one or more of these enzymes might add sugars other than glucose either directly to the glucose-accepting nucleotide or to the glucose of glucosylated DNA. In this case, the principles disclosed here could be applied equally well to polynucleotides containing these non- backbone, non-glucose sugars.

In addition, sugars other than glucose can be added. to glucosylated DNA by the action of one or more of the innumerable enzymes that are normally involved in producing or degrading oligo and poly -saccharides. For example, it has been reported (22) that both lactose synthetase and the galactosyl transferase subunit of lactose synthetase by itself can transfer galactose from the galactose donor molecule, UDP-galactose, to the disaccharide gentiobiose. (Gentiobiose is the disaccharide produced in vivo on the DNA of bacteriophase T6, a member of the T-even family of phages (21).) The enzyme, β-galactosidase, which normally is involved in removing galactose residues from oligo and polysaccharides, can also add, galactose to a wide variety of sugars (and other -OH containing compounds) either by condensation or by acting as a transferase (23). Thus, sugars other than glucose can be readily added to the glucose of glucosylated DNA by any of several enzymes that are not normally involved with nucleic acids. If galactose is covalently linked to DNA, the essential principle of this invention (the oxidation of non-backbone sugar moieties to produce aldehydes) can be performed enzymatically by the use of galactose oxidase (24) to generate an aldehyde group attached to the galactose ring. If only a single galactose were present, a monoaldehydic polynucleotide could be produced.

Thus, in general, the nucleic acid material used in performing this invention is not limited to glucosylated nucleic acids but also includes nucleic acids that contain non-backbone sugars other than glucose, nucleic acids containing carbohydrate chains, polyamino acid chains etc, or combinations thereof.

It is disclosed herewith that the glucose moiety of glucosylated DNA can be readily and specifically oxidized by periodate. The invention is not, of course, limited to the periodate technique.

It has been known for a long time that individual glucose molecules and glucose residues in carbohydrate chains are chemically reactive. For example, they can be oxidized by periodate to form dialdehydes provided that (unblocked) hydroxy groups are present on adjacent carbon atoms of the sugar ring (cis diols). Periodate oxidization of cis diols on sugar moieties has long been used to practical advantage. A recent application in the field of biotechnology involves the oxidization of the oligosaccharide chain attached to many species of monoclonal antibodies. A large variety of potentially useful entities have been attached covalently to the antibodies through the formation of adducts with the dialdehyde formed by periodate oxidation (6).

The present invention shows for the first time that the glucose on glucosylated DNA can be a site for the covalent attachment of a variety of entities to a polynucleotide chain. Although the glucose of glucosylated DNA has previously been covalently linked to compounds used to determine the presence of glucose on the DNA, this has been done only under conditions that destroy the polynucleotide chain or release the glucose from the base, hydroxymethylcytosine (see for example, reference 25). In short, this invention is based on the first and surprising realisation of a simple, non-destructive means of providing reactive attachment sites in large numbers on polynucleotides. Glucosylated DNA can be oxidized by periodate, as already indicated, and the oxidized product is also chemically reactive as shown by its ability to form stable adducts with a variety of chemical compounds. Molecules that have never before been covalently attached to DNA can be attached to glucosylated DNA that has been oxidised (DNA_ox).

The principles of this invention can readily be illustrated with any glucosylated DNA. In the following, the biologically produced DNA of bacteriophage T4, a member of the T-even family of bacteriophages, is used. Since essentially any DNA can be cloned (16,31,32) and amplified (16,31) in T4 genomes, it is now obviously straightforward to produce, in vivo, significant amounts of glucosylated DNA of any desired sequence. Thus, the results obtained with Bacteriophage T4 DNA can represent the kinds of results that one can achieve with any nucleic acid sequence whatsoever which contains a non-backbone sugar moiety. PREPARATION OF BACTERIOPHAGE T4 DNAs

DNA was isolated by standared methods of phenol extraction from purified T4 phage particles. The phage were purified by the well known technique of alternate cycles of high and low speed centrifugation.

T4 DNA that does not contain glucose was obtained by methods well known in the art. A multiple mutant phage stock (gene 56 amE10, gene 42 amN55, denA S112, denB Sa 9 and ale (TBI) ) was grown for one cycle of infection in the non-permissive host E. coli B^E and for a second cycle of infection in the non-suppressing host E. coli 834 as described (26). This DNA contains low levels of glucose because most of the HMC bases, to which the glucose is enzymatically attached, have been replaced by cytosine. This DNA will be referred to as T4 dC DNA or as non-glucosylated DNA.

If this same multiple mutant phage stock is grown in an amber suppressing host such as E. coli CR63, or B40 the T4 DNA will contain the unusual base HMC and have glucose residues attached to these bases. This T4 glucosylated DNA will be referred to as HMC DNA. It should be noted that these T4 dC and HMC DNA's are genetically identical but chemically different. Thus the different behavior of these two DNA's disclosed below can be explained only by these chemical differences.

HMC and dC DNA's were either sonicated five times for 30 seconds at maximum power with a micro tip in a Branson sonicator, model W185D, or digested with the restriction endonuclease Taql (Boehringer Mannheim) at 65° in T4 buffer (33mM Tris-Acetate, pH7.9; 66mM Potassium Acetate; 10mM Mg-Acetate; 0.5mM DTT and 100μg nuclease free BSA). Both of these treatments reduce the size of the DNA and the viscosity of the DNA solutions. The DNA solutions were then ethanol precipitated and resuspended in either water or 50mM NaCl at 0.2 to 2.0mg/ml. Wild type T4 phage, multiple mutant phage suitable for making isogenic glucosylated and non-glucosylated T4 DNA as well as suitable E. coli host strains can be readily obtained from any number of sources well known in the art.

PERIODATE TREATMENT OF DNA

To 10 volumes of sonicated or Taq I restriction endonuclease digested DNA, one volume of 0.5M Sodium Acetate, (adjusted to pH 5.6 with acetic acid) was added. Then, one volume of freshly prepared 0.6M Sodium Periodate dissolved in water, was added. The mixture was incubated at room temperature, in the dark for 30 min - 6 hours. Excess periodate was removed by dialysis, in the dark, first against 50mM Sodium Acetate, pH 5.6 and then against 50mM NaCl. The periodate treated DNA's were used immediately, stored at 4°C or frozen at 20°C until they were analysed or treated as described below.

Periodate oxidation of nucleic acids is not new. Previously, the periodate oxidation of the sugar moiety of RNA molecules, but not of DNA moecules has been reported in the scientific literature. The ribose moiety at the 3' end of tRNA molecules uncharged with amino acids provide a structure (a cis-diol) that has been the object of periodate oxidation to a dialdehyde since at least 1960 (27). These oxidized tRNA molecules have been reacted with a number of reagents. See (5) for a review of this literature up to 1979. RNA molecules carrying reporter groups attached to these periodate produced aldehydes have been used as probes to locate the cellular sites of complementary DNA sequences (28). In this case a fluorescent reporter molecule was used. However, with tRNA molecules, only a single fluorescent reporter molecule could be attached directly to each RNA molecule. To our knowledge, fluorescent molecules have not previously been directly attached to DNA molecules (9). With glucosylated DNA oxidized as just described (DNA_ox) not only can fluorescent molecules be directly attached to DNA, but multiple fluorescent molecules can be attached (see below).

The important fundamental structural difference between oxidizable RNAs and DNAs is that with RNA the oxidisable group (ribose) is an integral part of the sugar-phosphate backbone of the nucleic acid, while with glucosylated DNA the oxidizable group (glucose) is not a structural component of this backbone.

The present demonstration of the use of periodate to oxidize nucleic acids is new in at least five ways. The oxidizable group is glucose, not ribose. The oxidizable group can be located at any position along the polynucleic acid chain instead of only at the 3' end. The oxidization of a sugar moiety on a naturally occuring DNA is new. The production of more than a single pair of aldehydes on a polynucleotide chain (of either RNA or DNA) by oxidization of sugar moieties is new and the oxidization of a sugar moiety that is not part of the sugar-phosphate backbone of a polynucleotide chain is new. Of course it is understood that in preforming this invention that other, perhaps less preferred methods, could be employed to oxidize the (non-backbone) sugar moieties of nucleic acid molecules. For example, it is well-known that lead tetra acetate can oxidize glucose and other sugars (29). In addition it has already been mentioned that enzymes can be employed to oxidize some (non-backbone) sugars that might be covalently attached to nucleic acid molecules.

REACTION OF DNA WITH 2,4 - DINITROPHENYLHYDRAZINE

2,4-dinitrophenylhydrazine (2,4-DNPH) is a well known chemical reagent specific for aldehyde and ketone groups. This reagent has been used to detect the dialdehydes formed by periodate oxidation of the 3' ends of tRNA molecules (27). Essentially this procedure may be used to detect the aldehyde groups produced by periodate treatment of glucosylated DNA. The procedure may be as follows:

To 2 volumes of DNA are added 1 volume of 3.3% acetic acid, pH 3.5 (prepared by diluting a solution of 10% acetic acid, adjusted to pH 3.5 with NaOH) , 3 volumes of 2-methoxyethanol free of peroxides and 4 volumes of a 1.2% solution of 2,4-dinitrophenylhydrazine in 2-methoxyethanol. After incubation at 30° for 1-8 hours, 6 volumes of water were added and the excess reagent eliminated by extracting five times with ethyl acetate.

Aldehyde groups are detected by this method only with the glucosylated DNA that has been treated with periodate (table 1). In this experiment, 9-10 μg of glucosylated (HMC) and non-glucosylated (dC) T4 DNA's that had been exposed to sodium periodate and then dialyized, as described above, were reacted with 2,4-dinitrophenylhydrazine. As controls, 9-10 μg of HMC and dC DNA's that had not been exposed to periodate; were also mixed with 2,4-DNPH and processed identically. After removal of the unincorporated 2,4-DNPH with ethyl acetate, only one of the samples (the periodate treated glucosylated DNA) retained any visual evidence of the yellow color characteristic of the 2,4-DNPH reagent. Each sample was then ethanol precipitated and resuspended in 400 μl of TE (10mM Tris-HCl, pH 8.0; ImM EDTA) and its optical density at 260, 280 and 360 mμ determined.

The results given in Table 1 are given as the average number of 2,4-DNPH molecules attached to the DNA per HMC or cytosine base. These calculations assume that the molar extinction coefficient of the 2,4-DNPH radical at 360my (O.D. 360) is 21,000 (27) that T4 DNA is 34% G + C base pairs (i.e. that 17% of the bases are HMC or cytosine) and that the O.D. 260 of the DNA-bound 2,4-DNPH is small relative to the O.D. 260 of the DNA. (This method of calculation probably under estimates the true level of attachment of 2,4-DNPH to DNA_ox because (unbound) 2,4-DNPH has a significant O.D. at 260mμ.)

These data demonstrate that under these conditions only glucosylated (HMC) DNA can be oxidized to a form that can react significantly with 2-4, DNPH. The only chemical difference between the HMC and the dC DNAs is that the HMC DNA contains the glucose-accepting nucleotide HMC, and that glucose is attached to all of the HMC bases. Thus, it seems almost certain, from this data alone, that the periodate reacts specifically with the glucose moieties to form aldehydes and that the 2-4, DNPH reagent is attached specifically to these aldehydes. More generally this data establishes that numerous glucose residues on individual DNA molecules can be oxidized to a form that permits them to be chemically coupled to specific compounds.

of course it is understood that the exact conditions used for oxidation will depend on whether one wants to produce, in a single step DNA molecules having only one, a few, many or all of the glucose moieties oxidized. This will be determined by the particular application envisaged for these molecules. Since multiple sites for oxidation and hence adduct formation are available, one might want to perform repeated cycles of limited oxidation and adduct formation. Such a protocol might be useful for, for example, the sequential addition of more than one kind of entity to individual DNA molecules.

SCHIFF BASE FORMATION

REACTION OF OXIDIZED DNA (DNA_ox) WITH m-AMINOBENZOIC ACID

Glucosylated DNA oxidized by periodate can undergo a variety of additional chemical reactions. The reaction of DNA_ox with 2,4-DNPH has been presented above. The use of DNA_ox to form Schiff bases that can be reduced with cyanoborohyride is now described.

Schiff base formation can be demonstrated with a compound (m-aminobenzoic acid) that gives a fluorescent conjugate in DNA. This is the first time that a fluorescent group has been directly attached covalently to DNA (9). The ability to covalently link a large number of fluorescent entities directly onto individual DNA molecules has obvious potential practical applications in the field of DNA-based diagnostics. Fluorescent DNA is also convenient experimentally because it is easy to follow (see below).

Convenient conditions for producing Schiff bases with glucosylate DNA and their reduction by cyanoborohydride are 5 volumes of periodate treated DNA (at 100 to 2000 μg/ml in 50 mM NaCl). 1 volume of 10% acetic acid (adjusted to pH 3.5 with

NaOH). 2 volumes of 8mM m-aminobenzoic acid (Fluka) in 1% acetic acid (prepared by diluting the 10% solution with water) and 2 volumes of freshly prepared 8mM sodium cyanoborohydride (Fluka) in water. These reactions were carried out at room temperature for 10 minutes to 16 hours. Of course other conditions of temperature, buffers, pH, concentrations and types of reactants, and times of reaction can also result in the formation of Schiff bases with oxidized DNA and the reduction of these Schiff bases. The conditions and reactants used here are chosen merely to illustrate that DNA_ox is very reactive chemically under suitable conditions. In this case, DNA_ox is reacting with an NH₂ (amine) containing compound. Aromatic amine compounds are particularly useful for this purpose since they are more nucleophilic than aliphatic ones and, because of thier characteristic pK's of protonation, one can obtain a large measure of selectivity relative to aliphatic groups (30). Tritiated cyanoborohydride (Amersham TRK.708) was diluted 100-fold with unlabelled 3mM cyanoborohydride to a specific activity of 34 mCi/mM.

The input radioactivity for the experimental data presented in Table 2 was from 5.8-7.0 x 10⁴ cpm. The glucosylated and non-glucosylated DNA's are the same as those used for the experiment described in Table 1.

In this experiment, 10 μl reaction mixtures containing 1.85 μg of DNA were incubated for 30 minutes. The DNA was then acid precipitated, collected on glass fiber filters and counted in a toluene-based scintillation cocktail.

This data clearly show that DNA cannot accept ³H - from cyanoborohydride unless the DNA contains glucose and has been treated with periodate. That is, ³H - incorporation occurs only with DNA_ox. More generally, this result shows that adducts with DNA can be readily formed when DNA containing one or more cis-diols is oxidized to produce aldehydes.

The data in table 2 also shows that significant incorporation of the ³H - into DNA_ox does not occur if the amine, m-aminobenzoic acid is omitted. This result strongly suggests that Schiff bases are formed between DNA_ox and m-aminobenzoic acid and that the ³H - is incorporated into the DNA as the result of reducing these Schiff bases. That is, cyanoborohydride will reduce the Schiff bases much more readily that it will reduce directly the aldehydes produced by periodate oxidation. Visual examination of DNA pellets and agarose gel results discussed below strongly support the conclusion that these Schiff bases are essential intermediates for the efficient formation of the DNA_ox - m-aminobenzoic acid adducts.

The results of reacting m-aminobenzoic acid with glucosylated and nonglucosylated DNA's can also be seen visually. Solutions of m-aminobenzoic acid are mildly fluorescent (bluish) when viewed under ultraviolet illumination in the range of 200 nm to 360 nm. DNA_ox - adducts with m-aminobenzoic acid are also mildly fluorescent. Thus, when DNA_ox is mixed with m-aminobenzoic acid as described above and then ethanol precipitated by standard techniques, the pellet of DNA is strongly fluorescent. Visibly detectable levels of fluorescent DNA are produced only with DNA_ox. These results are summarised in Table 2.

Evidence that the fluoresence of m-aminobenzoic acid is intimately associated with DNA_ox was obtained by agarose gel eletrophoresis of aliquots of these various reaction mixtures or of the DNA recovered by ethanol precipitation as described below. UV illumination of the agarose gels, without prior staining with ethidium bromide (EtBr, a common stain for DNA) , revealed that flouresence co-migrated with the DNA only for the DNA_ox sample (data not shown). The location of the DNA samples in the agarose gel was determined after this initial examination by staining the gel with EtBr.

With m-aminobenzoic acid, visible evidence of adduct formation with DNA_ox was observed only if cyanoborohydride was added to the reaction mixture. This result strongly suggests that Schiff bases are essential intermediates for the efficient formation of the adducts. This result is confirmed by the gel analysis presented below.

The compound, m-aminobenzoic acid, used here to demonstrate the formation of adducts with DNA_ox that presumably involve Schiff base intermediates, has recently been shown to be very efficient in forming Schiff base adducts with aromatic-aldehyde derivatives of proteins (30). Many other compounds originally developed for attachment to protein-aldehydes thus can also be found to readily form stable adducts with DNA_ox (glucosylated DNA that has been oxidized). AGAROSE GEL ANALYSIS OF DNA_ox - ADDUCTS

The evidence presented thus far clearly indicates that m-Aminobenzoic acid can form significant levels of adducts with DNA only if the DNA is DNAox (i.e. glucosylated DNA that has been treated with periodate). Another way of demonstrating the formation of these adducts is presented in Figure 1.

In this experiment only glucosylated DNA was used. The DNA was first digested with the restriction endonuciease Taq 1 (as described above), phenol extracted, ethanol precipitated and resuspended in 50 mM NaCl. One portion of the digested DNA was oxidized with periodate as described above and then both oxidized and non-oxidized DNA's were mixed with m-aminobenzoic acid in the presence of cyanoborohydride as described above. Approximately 1 μg of DNA from these reactions and several control reactions were analysed by electrophoresis on a 1.2% agarose gel (Sigma type II, made and electrophoresized in TEB buffer (90mM tris; 90mM Boric Acid; 2.5 mM EDTA; pH 8.3). After electrophoresis the gel was stained with ethidium bromide and photographed by standard techniques. Also included on the gel are samples of the input DNA's (columns (a) and (b) are DNAox and columns (c) and (d) are glucosylated DNA that was not treated with periodate) and double stranded DNA size markers (columns (i) - (1): (i)4.3 and 3.6 kilobases (kb); (j) 5.4, 1.4 and 1.1 kb; (k) 4.0, 2.3 and 1.3 kb; (1) 3.2, 2.7 and 1.7 kb. The faint band near the top of column (1) is a partial digest product. Examination of the figure shows that the band pattern in each of the taq 1 digests (columns (a)-(h)) is similar. However, in comparing different columns, it can be seen that small, but significant differences occur in the extent of migration of individual DNA bands into the gel. These differences are reproducible. The most important comparisons are those involving DNA_ox (columns (e)-(g)). In particular, when DNA_ox reacted with both m-aminobenzoic acid and cyanoborohydride (column

(e) ), the individual bands migrate more slowly (and thus have a greater mass) than the individual bands from either of the other two reaction mixtures containing DNA_ox (compare column (e) to columns (f) and (g) ). The slower migration occurs only if both m-aminobenzoic acid and cyanoborohydride are present (column (e) ). If m-aminobenzoic acid is omitted from the reaction (column (f) ) or if cyanoborohydride is omitted from the reaction (column (g) ) the reaction products have the same mobilities as unreacted DNA_ox (columns (a) and (b) ).

In contrast, reaction of the control DNA (glucosylated DNA that had not been exposed to periodate) with both m-aminobenzoic acid and cyanoborohydride did not affect its mobility relative to the unreacted control DNA (compare column (h) to columns (b) and (c) ). Thus, once again, as was also demonstrated by the data of tables 1 and 2, the creation of new products is specific for DNA_ox (glucosylated DNA oxidized by periodate).

Comparison of columns (a) and (b) with columns (c) and (d) shows that the electrophoretic migration of individual bands of the oxidized and non-oxidized glucosylated DNA are slightly different. The reason for these differences is not known. Whatever the cause, it does not affect the primary conclusion that only DNA_ox molecules, when mixed with m-aminobenzoic acid in the presence of cyanoborohydride can react to product detectable amounts of DNA-containing products having reduced electrophoretic mobilities.

The reduced electrophoretic mobilities of individual DNA bands is thus another demonstration that numerous adducts can be readily formed chemically with individual DNA_ox molecules.

The molecular weight of m-aminobenzoic acid is 137.1. If one molecule of this compound was attached to a glucose on glucosylated DNA, the mass of that GC base pair (actually a G-HMC base pair since the glucose is attached to the HMC bases) would be increased by 12%.

If one m-aminobenzoic acid molecule were attached to every glucose on T4 DNA, the mass of an average T4 restriction fragment would increase by 5% (17% of T4 bases are HMC; the molecular weight (MW) of an A-T base pair is 653 and the MW of a C-HMC plus glucose plus m-aminobenzoic acid base pair is about 950). In fact, a graphic analysis of the observed mobilities of several pairs of corresponding bands in columns (e) and (g) agrees reasonably well with this theoretical calculation. That is, the apparent masses of selected bands in column (e) are between four and ten percent greater than the masses of the corresponding bands in column (g).

Significantly increased mass of individual molecules of DNA_ox and consequently significant reductions in electrophoretic mobilities could not occur unless m-aminobenzoic acid had been attached to a significant fraction of the oxidized glucose moieties. Thus, numerous m-aminobenzoic acid molecules must have been added to individual DNA_ox molecules in the experiment analysed in Figure 1. Furthermore, the reactions appear to have occured with nearly equal efficiencies on all or almost all of the molecules. If this were not the case, individual bands on the gel would be less sharp or well defined and would tend to smear into one another. Thus, the observation of sharp bands having reduced mobilities shows that numerous adducts can be readily formed with essentially all of the DNA_ox molecules and that the reaction producing these adducts can be quantitatively quite efficient.

An exactly similar analysis was done at the same time with a derivative of m-aminobenzoic acid that is capable of chelating Fe and Ga. Adducts between DNAox and this compound, [m-aminobenzoyl-ferrioxamine B (mF) ], show a greater mobility shift than obtained with m-aminobenzoic acid, consistent with the greater mass of mF (data not shown). However, in this case, the bands of DNAox-mF are less sharp than with m-aminobenzoic acid. Polynucleotides attached to complex compounds like mF can undoubtedly be produced much more easily by direct chemical coupling to polyaldehydic DNA than by incorporation of a bulky, complex nucleotide into a polynucleotide chain by either nick-translation or chemical synthesis. Thus, the use of polyaldehydic DNA allows one rapidly to synthesize novel polynucleotide derivatives and to test them for their behaviour under particular experimental conditions.

European Patent Specification No. 0243929 describes mF and related compounds which may be used in this invention.

in summary, the usefulness of glucosylated DNA has been demonstrated by the experiments disclosed. The glucose molecules attached to

glucosylated DNA can be oxidized by a simple chemical method to produce aldehyde groups which are highly reactive under appropriate conditions. The production of aldehyde groups on glucosylated DNA by periodate treatment was demonstrated by reacting the oxidized DNA with the aldehyde-specific reagent 2,4-dinitrophenylhydrazine. The reactivity of the aldehyde groups on DNA_ox was demonstrated by their reaction with 2,4-DNPH and by the formation of Schiff bases with m-aminobenzoic acid and the reduction of these Schiff bases with cyanoborohydride. The evidence presented suggests that the aldehydes so produced can be used to form stable adducts. The chemical methods used for these demonstrations are easy to perform and can readily be scaled up. Other kinds of adducts can also be formed.

It is clear from the results presented thus far that the aldehyde specific reagents can be covalently attached to either strand of the double stranded DNA-_ox molecules used to demonstrate this invention. Thus, it is obvious that reporter groups can also be added, with the same methodologies, to a double stranded nucleic acid having the .chemically reactive polyaldehydes on only one of the two strands of a duplex molecule, such as would be found in a probe-target hybrid.

NUCLEIC ACID HYBRIDIZATION STUDIES

The experiments presented thus far make it amply clear that one can oxidize non-backbone sugars to aldehydes and covalently attach aldehyde-reactive groups to the oxidized sugars.

The reactions are highly specific; significant attachement of the aldehyde reactive groups to DNA lacking oxidized sugars was not .observed.

However, it is not obvious from these results, how to best exploit these reactions in situations requiring nucleic acid hybridization. The data presented in tables 3 and 4 (and the subsequent data herein) indicate that the best approach for increasing sensitivity will be to attach the reporter group (e.g. biotin) to the probe after hybridization to the target. In these experiments, the hybridization efficiency of glucosylates DNA, polyaldehydic DNA (DNAox) and DNAox-adducts are compared under the relatively stringent hybridization conditions normally used for glucosylated DNA. It will be seen that the DNAox-reporter complexes hybridize with significantly reduced efficiency. This indicates that the use of probes made from DNAox-adducts in nucleic acid hybridization reactions, will encounter the same problems that limit the sensitivity and usefulness of other kinds of probe-reporter complexes. In contrast, DNAox molecules can hybridize with a normal or only moderately reduced efficiency. In any case, probes made from DNAox will encounter fewer of the problems referred to above than probes made from DNAox- reporter complexes and hence will allow the achievement of greater sensitivity than probes already linked to a reporter group.

Of course, one could also make probes of glucosylated DNA, hybridize the glucosylated probes to the target sequences and then oxidize the non-backbone sugar and add the reporter groups. This sequence of events would, require at least one additional manipulation by the end user. In addition, this approach would often not be desirable when the target nucleic acid is RNA because the riboses at the 3' ends of the RNA chains could also be oxidized (at least by periodate) and linked to the reporter groups.

The data in table 3 is from an experiment in which ³²P-RNA was hybridized for 17 hours at 66-67°C in 2 ml 2×SSC with polyaldehedic and wild-type (glucosylated) T4 DNA that had been loaded onto nitrocellulose filters as described (16), except that the T4 DNA was fixed to the filters by UV treatment (33) instead of baking at 80°C under vacuum.

Table 3

Sample amount of % hyb % hyb % hyb

# T4 DNA to to to per filter T4ox T4non-ox pBR filter

1 7. 5ug 14.6% 15.2% 0.6%

2 7. 5ug 14.5% 16.3% 0.5%

3 0.75ug 3.4% 3.5% 0.7%

4 0.75ug 2.8% 3.3% 0.7%

The ³²P-RNA (40,000 cpm in 2 μl; contains 327 bases of T4 and about 100 bases of non T4 DNA) was prepared as a Riboprobe, by standard proceedures, from a clone of T4 DNA coming from the gene 32 region (approximate kilobase co-ordinates of 146.25 to 146.5 on the standard T4 map). In each hybridization vial, done in duplicate, there were three filters; one was charged with 0.8 ug of plasmid pBR322 DNA and the other two were charged with equal amounts of either polyaldehydic or wild-type T4 DNA. After hybridization, the filters were batch washed at room temperature: first two times for at least 30 minutes in 2×SSC, then once for at least 15 minutes in

0.2×SSC and finally put in water for sorting. Thus, in a hybridization vial, the ³²P-RNA could hybridize with either polyaldehydic or glucosylated DNA. The slightly higher hybridization efficiency with the non-oxidized DNA probably reflect only experimental error. However, it could mean that the hybridization kinetics of RNA with DNAox are slightly slower, or that the thermal stability of RNA-DNAox hybrids is slightly less than for the controls. In any case, it is clear that DNAox can efficiently hybridize with complimentary nucleic acid sequences. As already indicated, this result is important for determining how to employ polyaldehydic DNA in DNA diagnostic tests. This highly efficient hybridization of polyaldehydic DNA would not have been predicted by molecular biologists.

We note that the aldehydes might interact with the NH₂ groups on adenine, guanine or cytosine bases and thus might, a priori, be expected to reduce hybridization efficiencies. In any case, this does not have to occur to a significant extent and consequently may have important practical uses. In another RNA-DNA hybridization experiment (table 4) increasing amounts of ^32P-RNA (1-10 μl of the same Riboprobe preparation as used above) were hybridized, in solution (66°C for 18 hours in 1 ml containing 800 μl of 2×SSC and 200 μl of the denatured DNA in essentially 4×SSC (16)), with a constant amount of either wild-type T4 DNA (12.9 ug) oτr polyaldehydic T4 DNA (11.0 ug) that had been heat denatured in alkali as described previously (16).

Table 4

amount of % RNase resistance 3²P-RNA after hybridization with T4ox T4non-ox

1 μl 38% 43%

3 μl 24% 25%

5 μl 17% 17%

10 μl 11% 13%

In this experiment, the extent of RNA-DNA hybrid formation was determined by measuring theϊ fraction of the input counts that became resistant to RNase digestion (to 100 μl of hybridization solution, 1 ug of Yeast tRNA and 5 ug of pancreatic RNase were added, incubated for 45 minutes at room temperature and acid precipitated).

Under these conditions, about 2% of the input. 3² P-RNA is RNase resistant.

Again, the results of this experiment show that the polyaldehydic DNA can hybridize to RNA with essentially normal efficiencies over a 10-fold range of RNA to DNA ratios. Further experiments have been performed which also demonstrate conclusively the potential of oxidized glucosylated DNA as a complimentary probe. It was found that aldehydic DNA hybridizes with 83% of the efficiency of the untreated DNA.

The following test was carried out under conditions in which the probe was present in non-saturating quantities relative to the immobilized target. Under these conditions, even a small impairment in capacity to hybridize will show up. On the other hand, probes are used in real life in saturating excesses over the target. Thus, the 83% efficiency of oxidized (i.e. aldehydic) T4 relative to that of untreated T4 DNA represents a difference that is regarded as quite insignificant under saturating conditions. These findings are surprising and very significant.

¹⁴C-labelled T4 DNA (wild type) was prepared from phage particles by standard methods. Samples of this DNA was oxidized as described above. The results are tabulated in table 5 below. The absolute efficiency of the assay (25, 15 %, see last line of the tabulation) is satisfactory, and therefore the relative efficiencies given in the first line of the tabulation can be taken to be valid.

The hybridization method was essentially that of Mattson et al. (1983) J. Mol. Biol. 170343-355, except that water was used in place of formamide and the hybridizations were done at 66°C instead of 42°C.

The following solutions were made up:

TSE

10mM Tris-HCl, pH 8.0 50mM EDTA, pH 8.0 500mM NaCl

Denhardt solution (fifty-fold strength)

5 g Ficoll

5 g Polyvinylpyrolidone

5 g BSA water to 500 ml

Neutralization mixture

100ml 1M HCl

20ml 1M Tris-HCl, pH8.0

80ml SSC (below), twenty-fold strength 20ml water

SSC

0.15M NaCl

0.015M Tri-Sodium Citrate

Each 1ml of the hybridization solution is prepared by first mixing 1-10 microlitres ¹⁴C-DNA solution (the actual colume is chosen so as to give the required number of counts) with 0.25 ml of solution TSE (see above) plus 0.25ml of 1M NaOH.

This mixture is placed in a boiling water bath for 10 minutes and then quenched in an ice bath. Neutralization mixture (0.55,1) is then added and the pH adjusted if necessary to about 7 with 1M NaOH. The total volume should now be 1.05ml. A sample (0.05ml) is withdrawn for counting (acid precipitation or spotting on GF/C glass-fibre filters) to leave a final volume of 1ml. Should 2ml or 3ml of solution be required, the quantities are adjusted accordingly. To each 1ml of the resulting solution was added 0.92ml water, 0.04ml of Denhardt's solution (fifty-fold strength) and 0.04ml heat-denatured E.coli DNA (1.0mg/ml, in water). The solution is then ready to be transferred to the hybridization vials (glass, siliconized). Each vial receives 2ml of the final mixture. Nitrocellulose filters were charged with an excess of single-stranded DNA according to the method of Mattson et al. 1983 (loc. cit.). Three types of filter were prepared: T4 DNA (unmodified) 8 microgrammes/ug/filter), calf thymus DNA (4 microgrammes/filter), and salmon-sperm DNA (4 microgrammes/filter). The filters were pre-treated in a mixture of 10 volumes of SSC (double strength) and 1 volume Denhardt solution (fifty-fold strength) for at least 2h. at 66° prior to hybridization.

More than one filter can be put into each hybridization vial. Before they were placed in the vials the filters were coded with a soft-lead pencil and wetted with SSC (six-fold strength). The coding is done so that the filters need not be washed separately, and the wetting is carried out in order to be able to reject those few filters that do not wet satisfactorily.

Hybridization took place overnight at 66° in a shaking water bath. The filters were then washed all together in copious quantities of SSC (double strength) at room temperature for 30 minutes. They were then rinsed once in water, transferred to scintillation vials, dried, and counted.

In contrast to the above, the following again demonstrates the adverse effect on hybridization of labels.

Hydroxymethylcytosyl T4 DNA (T4-HMC) was digested with the restriction enzyme Tag 1 and labelled with ³²P, by exchange of the 5' phosphate groups by serial reaction with bacterial alkaline phosphatase and T4 polynecleotide kinase. A portion of the labelled DNA was oxidized, and reacted with biotin amidocaproyl hydrazide (BioHZ) (Sigma Chemical) to produce a biotin derivative of T4 (T4-Bio). The efficiency of hybridization of these probes to excess filter-bound DNA was compared.

In preparing probes, T4-HMC was digested with Taq l (2.5 U/ug) at 65 C, precipitated with isopropanol and sodium acetate, and resuspended at 1 ug/ul in TE (TE: 10 mM Tris pH 8.0, 0.1 mM EDTA. A portion of this DNA was adjusted to 20 mM Tris with 1 M Tris pH 9.5, and the 5' phosphate groups removed by incubation at 37 C for 30 minutes with bacterial alkaline phosphatase. After the incubation, the enzyme was removed by extraction with phenol and chloroform-isoamyl alcohol (24:1), and precipitated as above. The DNA was resuspended at 1 ug/ul in TE. The 5' -dephosphorylated DNA was incubated at 37 C for 30 minutes with gamma [³²P]ATP and T4 polynucleotide kinase under standard conditions [36], precipitated as above, washed with 70% ethanol, and resuspended at 1 ug/ul in water. 15 ul of the 3²P-labelled T4-HMC was oxidized for 30 minutes with 60 mM sodium periodate and 50 mM sodium acetate pH 5.6. The oxidized DNA (T4-OX) was precipitated and washed as above, then resuspended to around 0.5 ug/ul in 1% acetic acid pH 3.5.

The T4-OX was reacted overnight with an equal volume of 27 mM BioHZ (dissolved in 50% acetonitrile, l%> acetic acid pH 3.5). After the reaction, the

T4-Bio was precipitated and washed, and suspended in TES (10 mM Tris pH 8.0, ImM EDTA, 1% SDS) .

Amersham Hybond-N nylon filters were wetted, soaked in 20 × SSC (Maniatis et al. 1982), and air-dried. Aqueous solutions at 1 ug/ul were prepared from Taq 1-digested T4-HMC and sheared salmon-sperm DNA. These were denatured by boiling for 5 minutes, and then 5 ul lots were deposited in dots on the prepared filters (as two applications of 2.5 ul) : sufficient 5 ug dots were prepared for hybridization to each probe in triplicate. The filters were air-dried, and the DNA was fixed to the filters by UV-irradiation. Filters carrying "no DNA" weres also prepared. The filters were prehybridized at 42 C overnight, in 5 ml of 6 × SSC, 50% deionized formamide, 50 ug/ul salmon-sperm DNA, 0.1% SDS, 2 × Denhardt's solution (50 × Denhardt's: 1% Ficoll, 1% polyvinylpyrrolidone, 1% bovine serum albumin).

In the hybridization experiments, the T4 DNA probes (T4-HMC and T4-Bio) were denatured by boiling for 5 minutes, and added to the tubes of filters. Hybridization was at 42 C for 8 hours, with constant agitation. The filters were rinsed at ambient temperature in 2 × SSC, then washed at 50 C for 15 minutes in 0.2 × SSC, 0.1% SDS. Filter-bound radioactivity was determined by Cerenkov counting.

Table 6 below shows the filter-bound counts expressed relative to (100% + amount of ³²P-T4-HMC bound to T4-HMC), after subtraction of background counts, and normalization to a constant number of counts added to each hybridization mix.

TABLE 6

WASH DNA ON FILTER

PROBE TEMP. T4-HMC SS-DNA NO DNA

50 100% 0.6% 0.2%

T4-HMC 65 88.7% 0.5% 0.2%

75 30.5% 0.3% 0.3%

50 21.6% 1.4% 0.6%

T4-Bio 65 19.5% 1.4% 0.6%

75 7.2% 1.1% 0.5%

The raw data from which the above figures was derived are shown in table 7. The recorded cpms are the means of triplicates. The figures in parentheses are normalized to 10⁶ cpm added to each hybridization (the actual cpm added were: T4-HMC 947400; T4-Bio 998100).

TABLE 7

PROBE TEMP. T4-HMC SS-DNA NO DNA 50 6632 (7000) 43 (45) 12 (13)

T4-HMC 65 5880 (6206) 31 (33) 12 (13)

75 2024 (2136) 21 (22) 22 (23) 50 1509 (1512) 100 (100) 43 (43)

T4-Bio 65 1362 (1365) 97 (97) 35 (35)

75 502 (503) 77 (77) 34 (34)

With respect to the effect of temperature on probe binding, the T4-Bio probe has a similar melting profile to T4-HMC, suggesting that reduced stability of the duplex is not responsible for the lower binding of the derivative.

Thus, in conclusion, the T4-Bio DNA hybridizes about one-fifth as efficiently as native T4-HMC (21.7%), and the T4-Bio derivative has a similar melting profile to T4-HMC, implying that the modified DNAs do not form less stable duplexes.

Thus, the data presented in Tables 6 and 7 illustrate that the hybridization efficiency is significantly reduced, to approximately 22%, when a commonly used reporter group, biotin, is attached to the DNA-OX probe molecules before the hybridization. The effect on hybridization, of any particular reporter group, is obviously determined both by the size and chemical nature of that group. Larger, more complex reporter groups added before hybridization would likely give even greater reductions in hybridization efficiencies. Clearly, some potentially useful reporter groups, such as a large, highly antigenic protein, would not be a useful reporter group if it were added to the probe prior to the hybridization. Thus there is a clear advantage to be able to add the reporter groups after hybridization.

It has been mentioned previously that there is a prior disclosure of biotin being added to glucosylated T4 DNA after periodate treatment (European Patent Specification No. 0133473). However, no hybridization data was presented. The data given here in Tables 3-7 clearly demonstrate two important points. First, adding a reporter group to glucosylated DNA before hybridization does not offer any obvious advantages, in the hybridization reaction, over other methods of preparing non-radioactive DNA probes. Second, and far more important for the present invention, these data show that adding reporter groups to glucosylated DNA before hybridization does not exploit the unique potential of this kind of DNA. The real potential of glucosylated DNA can be fully exploited only when it is realized that polyaldehydic DNA, all by itself, is the DNA probe of choice. Polyaldehydic DNA is the probe of choice because it can hybridize efficiently to target sequences, thus allowing virtually any reporter group, whatsoever, to be added to a probetarget hybrid molecule.

In accordance with the principles of this invention, the preferred method of practicing the invention in the field of DNA-based diagnostics would be to produce glucosylated DNA probe sequences in vivo, in E. coli by employing, as described in (31), Bacteriophage T4 (T4) denB mutant phage, in conjunction with plasmids containing both a region of homology with T4 and the (preprobe) sequence or sequences to be converted into glucosylated or polyaldehydic probes. Although the in vivo produced glucosylated probes can be of any length, it is preferred to introduce recognition sites for the restriction endonuclease Taq 1 (one of the few restriction endonucleases that can efficiently cut glucosylated DNA) at regular intervals into the preprobe sequence, say every 50-100 base pairs, in order to (1) facilitate the purification of the glucosylated probes away from the T4 sequences, by hybridization to complimentary single stranded sequences fixed to a solid support followed by elution of the now purified glucosylated probes, and (2) in order to have short probes that exhibit favourable hybridization kinetics with target sequences.

Attachment of reporter groups (e.g. biotin) to polyaldehydic probes after hybridization to the target sequences provides better opportunities to increase the signal to noise ratio. One could oxidize the glucosylated probes to produce polyaldehydic probes either before or after the hybridization step. It is greatly preferred, particularly where the target nucleic acid is RNA, to hybridize with polyaldehydic probes because oxidization after hybridization could reduce the signal to noise ratio, a consequence of the oxidization of the ribose moiety at the 3' end of the RNA target molecules. Clearly probes produced according to this invention allow a large degree of flexibility for the order in which the operations can be preformed. Thus, both the chemically reactive aldehyde groups and the reporter groups can be introduced either before or after the hybridization step.

Polyaldehydic probes can readily be employed with currently available detection systems such as avidin-biotin detection systems and immunological detection systems (i.e. an antigenic entity could be linked to the aldehyde group). Polyaldehydic probes also create the possibility of employing new detection systems, for example, a component of a light generating system, such as a luciferase-like enzyme, could be attached to the aldehyde groups.

As will be readily appreciated, the invention is not limited to the above exemplifying and descriptive matter but has very broad application. Modifications and alterations of the above within the competence of the skilled reader are thus included.

REFERENCES

1) Toulme, J.J., Krisch, H.M., Loreau, N., Thoung, N.T., and Helene, C. (1986). Specific inhibition of mRNA translation by complementary oligonucleotides covalently linked to intercalating agents. Proc.Natl.Acad. Sci. (USA) 83 : 1227-1231.

2) Dreyer, 6.B., and Dervan, P.B. (1985) Sequencespecific cleavage of single stranded DNA: oligodoxynucleotide. EDTA. FE(II). Proc.Natl.Acad. Sci. (USA) 82: 968-972.

3) Syvaner, A., Alanen, M., and Soderlund, H., (1985). A complex of single-strand binding protein and M13 DNA as hybridization probe. Nucl.Acid.Res. 13; 2789-2802.

4) Chan, V.T-W., Fleming, K.A., and McGee, J.O'D. (1985). Detection of subpicogram quantities of specific DNA sequences on blot hybridization with biotinylated probes. Nucl.Acid.Res. 13: 8083-8091.

5) Hansske, F., and Cramer, F. (1979). Modification of the 3' terminus of tRNA by periodate oxidation and subsequent reaction with hybrizides, Methods of Enzymology (Academic Press, Ed. Moldave, K., and Grossman, L.) Vol. LIX: 172-181.

6) Rodwell, J.D., and McKearn, T.J. (1985). Linker technology: antibody-mediated delivery systems. Biotechnology 3: 889-894.

7) Van Prooijen-Knegt, A.C., VanHoek, J.F.M., Bauman, J.G.J., Van Duijn, P., Wool, I.G., and

Van der-Ploeg, M. (1982). In situ hybridization of DNA sequences in human metaphase chromosomes visualized by an indirect fluorescent immunocytochemical procedure. Exp.Cell Res. 141: 397-407.

8) Tchen, P., Fuchs, P.P., Sage, E., and Leng, M. (1984). Chemically modified nucleic acid as immunodetectable probes in hybridization experiments. Proc.Natl.Acad.Sci. (USA) 81: 3466-3470.

9) Landegant, J.E., Jansen In De Wal, N., Baan, R.A., Hoeijmakers, J.H.J., and

Van der-Ploeg, M. (1984). 2-Acetylaminofluorene modified probes for the indirect hybridocytochemical detection of specific nucleic acid sequences. Exp.Cell Res. 153: 61-72.

10) Ziegler, K., Schmitz, I., Zahn, H., (1977). Introduction of new crosslinks into proteins. Advances in Experimental Medicine and Biology (Plenu, Press, Ed. Friedman, M.) 86A: 345-354.

11) Kornberg, S.R., Zimmerman, S.B., and Kornberg, A. (1961). Glucosylation of deoxyribonucleic acid by enzymes from bacteriophage-infected escherichia coli. J.Biol.Chem. 236: 1487-1493.

12) Revel, H.R., and Georgopoulos, C.P. (1969). Restriction of non-glucosylated T-even bacteriophages by prophage P1. Virology 39: 1-17.

13) Oligonucleotide synthesis a practical approach (1984) IRL Press (Oxford-Washington DC, Ed.

Gait, M.J.)

14) Casey, J., and Davidson, N. (1977). Rates of formation and thremal stabilities of RNA:DNA and DNA:RNA duplexes at high concentrations of formamide. Nucleic Acids Research 4: 1539-1552.

15) Roscoe, D.H., and Tucker, R.G. (1966). The biosynthesis of 5-hydroxymethyldoexyuridilic acid in bacteriophage-infected bacillus subtilis. Virology 29: 157-166.

16) Mattson, T., Van Houwe, G., and Epstein, E. (1983). Recombination between bacteriophage T4 and plasmid pBR322 molecules containing cloned T4 DNA. J.Mol.Biol. 170: 357-379.

17) Zimmerman, S.B., Kornberg, S.R., and Kornberg, A. (1962). Glucosylation of deoxyribonucleic acid II. α-glucosyl transferases from T4 and T6-infected Escherichia coli. J.Biol.Chem. 237: 512-518.

18) Josse, J., and Kornberg, A. (1962). Glucosylation of deoxyribonucleic acid III. α- and β- glucosyl transferases from T4-infected Escherichia coli. J.Biol.Chem. 237: 1968-1976.

19) Tomaschewski, J., Gram, H., Crabb, J.W., and Ruger, W. (1985). T4-induced α- and g-glucosyltransferases: cloning of the genes and a comparison of their products based on sequence data. Nucleic Acid Research 21: 7551-7568.

20) Gram, H. and Ruger, W. (1986). The α-glucosyltransferases of bacteriophages T2, T4 and T6. Mol. Gen. Genet. 202: 467-470.

21) Kuno, S., Lehman, I.R. (1962). Gentiobiose, a constituent of deoxyribonucleic acid from coliphage T6. J.Biol.Chem. 237: 1266-1270. 22 ) Schanbacher, F.L., and Ebner, K.E. (1970) Galactosyltransferase acceptor specificity of the lactose synthetase A protein. J.Biol.Chem. 245: 5057-5061.

23) Wallenfels, K., and Weil, R. (1972) β-Galactosidase. The Enzymes (Academic Press, Ed. Boyer, P.D.) Vol. VII: 658-670.

24) Morell, A.G., Vyn Den Hamer, C.J.A., and Scheinberg, I.H. (1966). Physical and chemical studies on ceruloplasmid IV. Preparation of radioactive, sialic acid-free ceruloplasmid labelled with tritium on terminal D-galactose residues. J.Biol.Chem. 241: 3745-3749.

25) Lehman, I.R., and Pratt, E.A. (1960). On the structure of the glucosylated hydroxymethylcytosine nucleotides of coliphages T2, T4, and T6. J.Biol.Chem. 235: 3254-3259.

26) Kutter, E., and Snyder, L. (1983). Preparation of cytosine-containing T4 phage in Bacteriophage T4 (American Society for Microbiology, Washington D.C., Ed. Mathews, C.K., Kutter, E.M., Mosig, G., and Berget, P.B.) pp56-57.

27) Monier, R., Stephenson, M.L., and Zamecnik, P.C. (1960). The preparation and some properties of a low molecular weight ribonucleic acid from baker's yeast. Biochem. Biophys. Acta 43: 1-8.

28) Bauman, J.G.J., Wiegant, J., and Van Duijn, P. (1981). Cytochemical hybridization with fluorochrome-labelled RNA. II. Applications. The Journal of Histochemistry and Cytochemistry 29.: 238-246 .

29) March, J., (1968). Advanced organic chemistry: reactions, mechanisms, and structure. (McGraw-Hill) pp. 867-870.

30) Offord, R.S., and Rose, K.R. in Proteids of the biological fluids 34: (Pergamon Press, Peeters, H., ed.) (1986).

31) Mattson, T.L., and Epstein, E. (1983). Inviable phages, their production and DNA thereof.

PCT No. WHO 85/01297.

32) Casna, N.J., and Shub, D.A. (1982).

Bacteriophage T4 as a generalized DNA cloning vehicle. Gene 18: 297-307.

33) Khandjian, E.W. (1987). Optimized Hybridization of DNA Blotted and Fixed to Nitrocellulose and Nylon Membranes. Biotechnology 5: 165-167.

34) Styer, L. (1868) Fluorescence spectroscopy of proteins Science 162: 526-533.

35) Turcotte, A. , Gicquaud, C. , Gendreau, M. , St. -Pierre, S. (1984) Separation of virotoxins from the fungus Aminita virosa and comparative study of their interaction with actin in vitro. Can. J. Biochem. Cell Biol. 62 : 1327-1334.

36 ) Mariatis , Fritsch , and Sanbrook (1932 ) ; "Molecular Cloning - A Laboratory Manual " (Cold Spring Harbor Laboratory) . 37) Langer, P.R., Waldrap, A.A., and Ward, D.C. (1981) PNAS USA 78 6633-6637.

38) Leary J.J., Brigati, D.J., and Ward, D.C. (1983) PNAS USA 80 4045-4049.

Claims

1. A polynucleotide probe having a base sequence hybridizable to a preselected polynucleotide and also having a multiplicity of reporter groups each linked thereto via an aldehyde group which has been provided by oxidation of a non-backbone sugar residue.

2. A polynucleotide probe as claimed in claim 1 which is a DNA.

3. A polynucleotide probe as claimed in claim 1 which is a RNA.

4. A polynucleotide probe as claimed in claim 2 or claim 3 which contains the base 5'-hydroxymethyl-cytosine.

5. A polynucleotide probe as claimed in claim 4 which is DNA from a T-even bacteriophage.

6. A polynucleotide probe as claimed in claim 4 which is a DNA in which 5' -hydroxymethylcytosine has been incorporated in place of some or all cytosine residues by a chemical, enzymatic, or biological synthetic technique.

7. A polynucleotide probe as claimed in claim 4 which is a RNA into which 5' -hydroxymethylcytosine has been incorporated enzymatically or chemically.

8. A polynucleotide probe as claimed in any one of claims 1 to 7, wherein the non-backbone sugar residue is glucose.

9. A polynucleotide probe as claimed in claim 8, wherein the glucose residue has been oxidized by periodate oxidation.

10. A polynucleotide probe as claimed in any one of claims 1 to 7, wherein the non-backbone sugar residue is galactose and has been oxidised using galactose oxidase.

11. A polynucleotide probe as claimed in any one of claims 1 to 10, wherein the reporter group(s) is/are

(a) fluororescent or radiodetectable e.g. include an m-aminobenzoic acid residue, or (b) include biotin.

12. A probe as claimed in any one of claims 1 to 11 which is a diagnostic probe.

13. A process for preparing a polynucleotide probe which comprises providing a polynucleotide having a base sequence hybridizable to a preselected polynucleotide and also in which there are non-backbone sugar residues and oxidising the said residues to provide aldehyde groups.

14. A process as claimed in claim 13 and further defined by the specific feature(s) of any one or more of claims 2 to 10.

15. A non-T4 derived polyaldehydic polynucleotide having aldehyde groups provided by at least one oxidized non-backbone sugar residue.

16. A method of diagnosis not being one practised on the human or animal body and in which a probe as defined in claim 12 is employed to detect by hybridization a polynucleotide characteristic of a disease or disorder or stage thereof.

17. A method of diagnosis in which a polynucleotide having aldehyde groups provided by at least one oxidized non-backbone sugar residue and which is hybridizable to a polynucleotide characteristic of a disease or disorder or stage thereof is subjected to hybridization with a sample suspected to contain said characteristic polynucleotide, and any hybrids thus-formed are labelled by attachment of reporter groups to said aldehyde groups and thereafter detected using said reporter groups.

18. A method of diagnosis in which a polynucleotide probe having non-backbone sugar residues and being hybridizable to a polynucleotide sequence characteristic of a disease or disorder or stage thereof is subjected to hybridization with a sample suspected. to contain said characteristic polynucleotide, after hybridization aldehyde groups are provided on the probe by oxidizing said non-backbone sugar residues, and any hybrids which have been formed are labelled by attachment of reporter groups to said aldehyde groups and thereafter detected using said reporter groups.

19. A method of diagnosis in which a polynucleotide probe which is hybridizable to a polynucleotide characteristic of a disease or disorder or stage thereof is subjected to hybridization with a sample suspected to contain said characteristic polynucleotide, said probe is thereafter provided with at least one non-backbone sugar residue and said residue(s) oxidized to provide aldehyde groups, and any hybrids which have been formed are detected by attachment of reporter groups to said aldehyde groups

20. A method of analysis in which a probe as defined in any one of claims 1 to 12 is employed to detect by hybridization the presence of a particular polynucleotide.

21. A method of analysis in which a polynucleotide having aldehyde groups provided by at least one oxidized non-backbone sugar residue and which is hybridizable to a polynucleotide which it is desired to detect is subjected to hybridization with a sample suspected to contain said polynucleotide to be detected, and any hybrids thus-formed are labelled by attachment of reporter groups to said aldehyde groups and thereafter detected using said reporter groups.

22. A method of analysis in which polynucleotide probe having non-backbone sugar residues and being hybridizable to a polynucleotide which it is desired to detect is subjected to hybridization with a sample suspected to contain said polynucleotide to be detected, after hybridization aldehyde groups are provided on the probe by oxidizing said non-backbone sugar residues, and any hybrids which have been formed are labelled by attachment of reporter groups to said aldehyde groups and thereafter detected using said reporter groups;

23. A method of analysis in which a polynucleotide probe which is hybridizable to a polynucleotide which it is desired to detect is subjected to hybridization with a sample suspected to contain said polynucleotide to be detected, said probe is thereafter provided with at least one non-backbone sugar residue and said residue(s) oxidized to provide aldehyde groups, and any hybrids which have been formed are detected by attachment of reporter groups to said aldehyde groups.

24. The use of a polyaldehydic polynucleotide as claimed in claim 15, or as defined in claim 15 but from a T4 origin, as a support for an immobilized enzyme, as a protein cross-linking agent, or as a aldehydic linker, e.g. wherein the polynucleotide is employed to link a toxin to a monoclonal antibody or to link a molecule having a biological effect on target cells to a monoclonal antibody for those cells

25. A duplex comprising a polynucleotide hybridized to a polynucleotide probe designed to detect a chosen base sequence or polynucleotide, the probe carrying a multiplicity of aldehyde groups or a multiplicity of reporter groups each linked to .the probe by an aldehyde group, and, in either case, the aldehyde groups each being provided by oxidation of a non-backbone sugar residue.

26. A duplex comprising a polynucleotide hybridized to a polynucleotide probe designed to detect a chosen base sequence or polynucleotide, the probe carrying intact non-backbone sugar residues so as to provide means for identifying the duplex by oxidizing said sugar residues to provide aldehyde groups to which reporter groups may be attached.