General Properties Of Plasmids

Plasmids are closed circular DNA molecules which replicate independently of the bacterial chromosome. They code for functions involved in their own life cycles and also for functions which affect the physiology of the host cell. Properties vary from plasmid to plasmid.

Maintenance is the ability of a plasmid to replicate in phase with host chromosome, hence to survive in host bacterium. Host range varies widely, some plasmids are restricted to a few closely related bacteria (e.g F factor of E. coli and related enterobacteria) others have a wide host range (eg. P-type plasmids can live in almost all gram negative genera tested).

Copy Number varies from 1 or 2 per chromosome (eg. F, P-types) to 40 or 50 (eg. ColEl). Number of copies has major effect on expression of plasmid-borne characters especially antibiotic resistance.

Transferability-Many plasmids can transfer themselves from host to host (eg. F & P-types, most R-factors). In addition some (eg. F) can also mobilize the chromosome. Others cannot self-transfer (eg. ColEl).

Mobilizability-ColEl cannot transfer itself, but it can be mobilized by F or ColV. Not all non-self-transmissible plasmids can be mobilized.

Incompatibility-Two plasmids which belong to the same family cannot coexist in the same cell. Such incompatibility groups are designated by letters of alphabet, eg. F-type plasmids include F, ColV, Rl; P-types include RPl and R751. Plasmids of a given incompatibility group are homologous in the DNA sequences involved in replication, maintenance etc. although the genes encoding other properties may be very different.

Size varies enormously. The F-factor is about 1% size of E. coli chromosome, whereas most high copy plasmids are much smaller (eg. ColEl about 1/10 size of F). Very large plasmids are sometimes found (eg. >10% of chromosome size) but are difficult to work with and few have been properly characterized.

Cryptic plasmids confer no identifiable phenotype on the host cell. The properties of these cryptic plasmids are presumably of some use to the host cell but are still unknown.

Most R-factors are of moderate to large size and present in 1-2 copies per chromosome. Most are self-transmissable at a low frequency, although derepressed mutants showing high transfer frequency are sometimes found. The original F-factor is such a "naturally-occurring mutant". R-factors belong to a wide range of incompatibility groups. Many carry one or many resistances to both antibiotics and/or toxic heavy metals and may also carry genes for colicins, virulence factors etc. Examples of plasmid types:

F-family. F itself carries no resistance markers and is a transfer derepressed mutant. ColV and ColVB colicin factors are in F-family. R100 codes for resistance to streptomycin, chloramphenicol, tetracycline, and sulfonamides; R1 to beta-lactams and kanamycin/neomycin. Hly (hemolysin) and Ent (enterotoxin) plasmids are usually F-type. ColV factors often specify virulence factors in addition to colicin V. F-types are self transmissable, large, single copy and can mediate transfer of many non-self-transmissable plasmids e.g. ColEl.

I-family. Many R-factors (e.g. R64, R144) and ColIa and ColIb factors. Some I-types carry both antibiotic resistance and colicin determinants. Self-transmissible, can mobilize other plasmids, large, single copy.

P-family. Noted for wide host range. Can self-transfer to almost every gram-negative genus tested in contrast to most plasmids which are restricted to a cluster of closely related bacteria. RPI (same as RP4, RK2), 62Md, single to five copies depending on host, beta-lactam, neomycin/kanamycin and tetracycline resistant is widely used in research. R751 specifies sulfonamide and trimethoprim resistance.

X-family. Unusual in being both multicopy (10-15 per cell) and self-transmissable. e.g. R6K, 26Md, coding for resistance to beta-lactams amd streptomycin.

Ecology

R-factors were in existence before therapeutic use of antibiotics but have become much more frequent in distribution since widescale use of antibiotics started. R-factors discovered first in Japan in dysentery causing Shigella. Shown to be transferred to and from intestinal E. coli. Sulfonamide (Sul) resistance found first, followed rapidly by tetracycline (Tet) and chloramphenicol (Cam) and streptomycin (Str).

by 1952 - 80% of Shigella SulR

by 1960 - 11% of Shigella SulR CamR TetR StrR

by 1969 - 34% of Shigella SulR CamR TetR StrR

One recent outbreak of infant diarrhoea caused by enteropathogenic E. coli harbored a plasmid specifying resistance to beta-lactams, SM, CM, TC, erythromycin, neomycin/kanamycin (NK) and novobiocin and was eventually killed with gentamycin (GM). In absence of antibiotic selection R+ bacteria tend gradually to decrease again in frequency. Major factor in R-factor spread is practice of feeding animals (e.g. pigs and chickens) antibiotics to increase yield. Recently some countries have banned use of human antibiotics in animal feed and there has been a major decline in frequency of R+ bacteria carried by farm animals.

Summary of Possible Mechanisms of Resistance:

The mechanism of R-factor encoded resistance is usually quite distinct from mechanisms observed in chromosomal mutants.

Antibiotic Chromosomal Resistance Plasmid Resistance

Beta-Lactams

Beta-Lactamase opens lactam ring producing penicilloic acid from penicillins and cephalosporanic acid from cephalosporins. Cephalosporanic acid is unstable and decomposes spontaneously to complex products. Most b-lactamases prefer either penicillins or cephalosporins, though a few attack both equally well.

Gram positive beta-lactamases are usually inducible. Hence a pronounced inoculum effect. If a small number of organisms is hit with a high concentration of antibiotic death occurs before sufficient beta-lactamase is produced to protect bacteria. If inoculum is heavy or initial antibiotic level is relatively low, bacteria survive as basal b-lactamase level rises to as much as 3% of total cell protein.

Gram negative beta-lactamases fall into two groups:

1) chromosomal; usually inducible, of low activity, prefer cephalosporins

2) plasmid encoded; constitutive, high activity, broad spectrum

Location is in periplasmic space - very efficient since only beta-lactams entering the PP-space can attack peptidoglycan synthesis. Chromosomal ampC mutants of E. coli which overproduce beta-lactamase increase resistance level from about 1-2 mg/ml of ampicillin to around 10 mg/ml. Compare this with RPl which confers a resistance level of several mg/ml. Certain Pseudomonads carrying RPl can use ampicillin as sole carbon source!

Methicillin and cloxacillin have bulky side groups and are resistant to almost all b-lactamases. They work well against resistant gram positives but fail to penetrate the gram-negative outer membrane. More recently, modified cephalosporins have been made which are both beta-lactamase resistant and penetrate the OM reasonably well eg. cefoxitin, cefuroxime. Alternatively can administer a mixture of e.g. ampicillin plus a lactamase inhibitor such as a clavulanic acid derivative. Recently, several new classes of beta-lactam have been discovered. Although the parent molecules are not of practical use, derivatives are being prepared and screened.

Thienamycins -- sulfur in ring replaced by carbon, sulfur found in side chain instead, wide spectrum of activity.

Nocardicins/ Sulfazecins -- lactam ring not fused to another ring at all, effective against gram negatives but no activity against Staph etc.

Theorigin of beta-lactamases is uncertain but they were around before antibiotic therapy. Presumably beta-lactamases were selected in environments where beta-lactam producing fungi compete with bacteria for survival. Possibly they are mutant versions either of: a) transpeptidases which gained the ability to destroy the substrate analogs - the beta-lactams or b) biosynthetic enzymes for beta-lactam synthesis in the original producer organisms.

Aminoglycosides

Resistance is due to modification by phosphorylation of -OH, adenylation of -OH or acetylation of -NH2 groups. Important enzymes are:

Streptomycin phosphotransferase

Neomycin/Kanamyin phosphotransferase

Streptomycin/Spectinomycin adenyltransferase

Gentamycin/ Tobramycin/Kanamycin adenyltransferase

Aminoglycoside acetyl transferases (3 enzymes, donate acetyl group to different positions) work on Neo, Kan, Gen and Tob to varying degrees.

Modified aminoglycosides can no longer induce the polyamine transport system by which they enter the cell and in addition they no longer inhibit their ribosomal target sites. Many different aminoglycosides and corresponding wide range of enzymes. Amikacin is a derivative of Kanamyin A with the l-NH2 group substituted with hydroxybutyrate. Amikacin is resistant to all modifying enzymes except one of the N-acetyl transferases and a recently discovered phosphotransferase specific for amikacin!

Origin of aminoglycoside modifying enzymes probably from Streptomyces strains which are the producer organisms, since these strains contain modifying enzymes. Role of such enzymes in producer is:

1. catalyse steps in biosynthesis
2. self-protection.

Chloramphenicol

Chloramphenicol modified by chloramphenicol acetyl transferase (CAT) in both gram positives and gram negatives. Replacement of the terminal -OH of chloramphenicol with fluorine results in non-modifiable yet still antibacterial derivatives. Like beta-lactamase, CAT is inducible in Staphylococcus but constitutive in E. coli. Induction process is slow due to fact that CM is inhibiting protein synthesis conjointly with inducing synthesis of CAT. Rapid induction can be achieved with the gratuitous inducer 3-deoxyCM which induces CAT but is not an enzyme substrate nor an inhibitor of protein synthesis.

Gram positive CAT enzymes are highly homologous with each other and so are gram negative enzymes. The two groups differ greatly from each other except for sequence homology in the CM-binding region.

Macrolides (Erythromycin & Lincomycin)

In gram positives, an inducible plasmid coded rRNA methylase modifies the 23sRNA of the 50s ribosomal subunit by dimethylation of a specific adenine. The modified ribosomes are resistant to erythromycin and lincomycin. Gram negatives are inherently resistant to macrolides which cannot penetrate the outer membrane effectively.

Tetracyclines

Tetracyclines are taken up by an energy-dependent transport system by bacteria but not eukaryotes. Since there is no similarity between Tet and any known transportable nutrients, the purpose of the transport system and its mechanism of operation are still baffling.

Plasmid coded resistance is similar in operation in both gram positives and gram negatives. Tet -resistance is typically two-level. A basal constitutive level of resistance protects by 5-10 fold relative to sensitive bacteria. In addition, exposure to Tet induces a second higher resistance level. Both resistance levels are due to successive drops in Tet accumulation. Tet induces production of proteins that are found in cell envelope and prevent uptake, somehow. In fact, Tet is actively deported from bacteria.

Lipophilic tetracylines e.g. monocycline probably penetrate membranes by diffusion well enough to short circuit usual uptake system -- and also avoid the waiting resistance proteins. Tet resistance plasmids therfore show only partial resistance to minocycline. Chlorotetracycline is intermediate in properties.

Sulfonamides and Trimethoprim

Sulfonamide inhibits dihydropteroate synthtase.

Trimethoprim inhibits dihydrofolate reductase.

Plasmid mediated resistance in both cases involved synthesis of a plasmid encoded alternative enzyme which is resistant to the inhibitor. R-factor dihydropteroate synthetase has same Km for p-aminobenzoic acid as chromosomal enzyme but it is several thousand times (Type I) or almost totally (Type II) resistant to trimethoprim. Sulfonamide plus trimethoprim is used for double blockade of folate pathway, however, R-factors carrying both sulfonamide and trimethoprim resistance now occur (e. g. R388, a W-type plasmid).

Transposons (Tn)

Mobile genetic elements which can jump between plasmids, phages, chromosome etc. Responsible for rapid dissemination of resistance genes many of which occur on plasmids as part of a transposon.

Tn 1,2,3 beta-lactamase
Tn4 beta-lactamase
Tn5,6 neomycin/kanamycin phosphotransferase
Tn7 trimethoprim
Tn9 chloramphenicol
Tn10 tetracycline

Bacteriocins are protein antibiotics sythesized by certain strains of bacteria and lethal to other, related but sensitive strains. Bacteriocins made by E. coli are called colicins, those made by Enterobacter cloacae are cloacins, etc. {Since most work is done on E. coli, bacteriocins from other bacteria are often referred to as colicins even though not strictly correct.} Antagonism between bacteria may be due to production of lactic acid, NH3, fatty acids, nucleotide analogs, etc. and is therefore not always due to colicin production. Colicins are plasmid specified.

Some Representative Colicins:

Plasmid Copy Colicin Axial

Plasmid Tra MW (Md) Number MW (Kd) Ratio Receptor

Colicin plasmids

Colicin Production

In a population of colicinogenic bacteria most cells do not produce colicin. At any given time around 1/1000 of the cells produces colicin. This is lethal for the producing cells. Colicin production is thus a sort of molecular kamikaze attack against sensitive bacteria. But note that it is not the colicin itself which kills the producer cell. Colicin negative mutants die when induced. Some other property which resides on the Col factor is lethal. Treatment with UV light, mitomycin C or other agents damaging the host cell DNA induce colicin production by most cells in a Col+ population. These are the same sort of treatments which induce lytic growth of many lysogenic phages such as lambda and f80. Note that colicin production does not require replication of Col plasmid DNA whereas phage production does require replication of phage DNA. Some colicins, e.g. colV tend to remain attached to surface of producing strain whereas others e.g. colicins E, are released as freely soluble proteins.

Colicin Receptors and Entry

Sensitive cells carry receptors for colicins which are outer membrane proteins, e.g. Bfe-protein is the receptor for vitamin B12, phage BF23 and colicins E. {See notes on outer membrane}. A single hit by one colicin molecule is sufficient to kill a sensitive bacterium. However, for every colicin that enters successfully, many others just bind to cell envelope. Colicin entry only occurs when receptor is in a particular position within the outer membrane which allows it to transfer colicin to inner membrane-perhaps the Bayers patches or TonB adhesion sites.

Loss of receptors or development of tolerance protects against colicins but interferes with outer membrane transport systems and hence is counter-selected strongly. Osmotically shocked cells are often sensitive to colicins even in the absence of appropriate receptors since outer membrane damage allows entry - this has been shown for E3 and M colicins.

Two classes of colicins from receptor-point of view:

Type A (A, E1, E2, E3, K, L) are inactive against tolA or tolB mutants but kill tonB mutants. I.e. need TolAB protein for uptake.

Type B (B, Ia, Ib, V, D, M) are inactive against mutants in the tonB gene, but kill tolA or B mutants. I.e. need TonB protein for uptake.

This classification is independent of the mode of action of the colicins and reflects two types of uptake system. Type A colicins probably enter via Bayer sites in a tolA,B dependent process. The TonB system carries out energy dependent uptake of colicins of group B and certain nutrients. The TonB protein provides energy for uptake, probably at the region of contact of inner and outer membranes. Note that many colicins are long thin molecules (axial ratios 10-20) and are long enough to penetrate the inner membrane while still attached to outer membrane receptor. This is true for colicins A, D, E1, I and K, all of which exert their killing effect on the inner membrane. In contrast, colicins E2 and E3 are shorter (axial ratio around 5) and act inside the cytoplasm after completely transiting the envelope. Thus membrane active colicins probably kill the cell while their outer end is still attached to the outer membrane. In fact, colicin E1 can kill cells while bound to sephadex resin whereas colicins E2 and E3 cannot. The N-terminus of colE1 binds to receptor while the C-terminus is the lethal end. Treatment of inhibited cells with trypsin reverses effect of E1, K or Ia indicating that the colicin molecule remains exposed at the cell surface.

Mode of Action

There are two major classes with respect to mode of action:

1) Intracellular target, e.g. DNA or ribosomal RNA.

2) Membrane target; usually inner membrane, rarely peptidoglycan.

Target Location

Receptor Class Membrane Cytoplasm Unknown

A (TolAB) E1, A, K, L E2, E3, DF13

B (TonB) Ia, Ib, M, B, V, D

Pesticin A1122

Cytoplasmic Acting Colicins:

a) E3 and DF13 are ribonucleases which cleave the 16s rRNA of the 30S robosomal subunit, releasing a fragment of 49 nucleotides from the 3'-terminus. This inactivates protein synthesis. The E3.I3 and DF13.IDF13 complexes are inactive against ribosomes (I3 & IDF3 are the immunity proteins - see below).

b) E2 is an endonuclease which destoys target cell DNA. Again, E2.I2 complex is enzymatically inactive.

For all of E2, E3 and DF13 mild proteolysis gives a C-terminal fragment, about 25% of the colicin, which possesses nuclease activity, binds immunity protein and is very basic (for binding to negatively charged nucleic acids). The N-terminal 25% is hydrophobic and probably involved in translocation across membrane. The central region of these colicin molecules interacts with the outer membrane receptor.

E2 and E3 are very similar in the N-terminal 75% of molecule - they share the same receptor. They differ in the C-terminus and have different nuclease specificities and immunity proteins. Although their respective immunity proteins (I2 and I3) are both very acidic, and of the same approximate size, they show no obvious structural relatedness.

Membrane Active Colicins.

ColE1, A, Ia, Ib and K form ion-permeable channels in cytoplasmic membrane or in artificial lipid vesicles. Primary event is collapse of proton motive force. Many secondary effects, e.g. drop in ATP levels, loss of K+ and Mg2+ from cell, cessation of macromolecular synthesis, etc. The cell membrane is not destroyed and most solutes do not escape. The colicin forms an ion-specific channel for protons and K+. A single colicin molecule is sufficient to depolarize a whole cell in a few minutes. Membrane active colicins have long thin molecules. The 18Kd C-terminal fragment of ColE1 is enriched in nonpolar aminoacids and is effective in depolarizing membrane vesicles or liposomes. The N-terminal fragment (40kd) contains receptor recognition region and is necessary only for penetrating the cell envelope.

Colicin M and Pesticin A1122 destroy the peptidoglycan rather than depolarizing the cytoplasmic membrane. These colicins need penetrate only as far as outer surface of cytoplasmic membrane, i.e. to site of peptidoglycan assembly. Cell lysis results in medium of low osmotic pressure whereas spheroplasts are formed at high OP. Pesticin A1122 is made by Yersinia pestis and kills Y. pseudotuberculosis, Y. enterocolitica, non-pesticinogenic Y .pestis and many E. coli but not E. coli K12. Pesticin hydrolyses the b-1,4 bond between N-acetyl glucosamine and N-acetyl muramic acid in peptidiglycan. Peptidoglycan from resistant mutants can be degraded, so resistance is due to inability of pesticin to penetrate OM.

Megacin A216 is a phospholipase made by Bacillus megaterium which hydrolyzes phosphatidylcholine to lyso-PC. Meg+ cells are protected by an immunity protein, which is specific for megacin and does not protect against the action of any other phospholipases.

Bacteriophage Related Bacteriocins

The R-type pyocins made by Pseudomonas aeruginosa are very different from other bacteriocins:

1) They are chromosomally encoded

2) The pyocins are similar in structure to a contractile bacteriophage tail with sheath, core, and fibers.

3) They contain over 20 different proteins

Several P. aeruginosa strains produce phages that show immunological cross reactions to R-pyocins. Furthermore, isolated tails from such phages show pyocin actovity. Thus pyocins are the remains of lysogenic phages rather than "true" colicins. Pyocin action is against cytoplasmic membrane. Marcescin A (from Serratia marcescens) is also a phage tail structure as are a few bacteriocins from bacteria, such as Clostridium and Vibrio. Few have been well characterized.

Colicin Immunity

Col plasmids specity immunity to the colicins which they produce. Hence a Col+ cell is immune to colicins produced by other Col+ cells of the same type. Immunity overload may occur if a massive dose of colicin is applied to an immune cell. Immunity is due to synthesis of immunity protein which binds and inactivates colicin. If too much colicin for immunity protein the cell is killed. Note that, unlike colicin which is only produced by a few members of Col+ population, the immunity protein is continually produced by all Col+ bacteria (Obviously - otherwise they would be dead!).

For membrane active colicins (E1, Ia, Ib, K, A), immunity is due to a plasmid coded inner membrane protein. For example, the Ia immunity protein of MW 14,500 protects membranes against colicin Ia but not against the closely related colicin Ib even though Ia and Ib share the same receptor, have same mode of action, and have extensive sequence homology. Thus cells immune to Ia can be depolarized by Ib or E1 but not by Ia (and of course vice versa).

Cytoplasmic acting colicins (E2, E3, CloDF13) - upon synthesis these colicins react with corresponding cytoplasmic immunity protein to give inactive complex. The colicin/immune protein complex is released eventually. How colicin is activated is uncertain - probably the immunity protein is lost during attachment to receptor of target bacterium. This not yet confirmed experimentally, although colicin/immunity protein complexes have been purified from culture fluids. Such complexes kill sensitive cells but are inactive against the biochemical target in vitro.

Colicin Ecology

Purpose presumably to kill related bacteria which compete for same ecological niche. Difficult to demonstrate experimentally though certain mixed infection experiments have indicated that Col+ bacteria can destroy colicin sensitive strains. Development of resistance to colicins is counterselected by loss of outer membrane transport ability.

Toxins are proteins which damage eukaryotic cells in contrast to colicins which kill other prokaryotes. Same basic principle of delivery system plus kill system on same molecule. However, if the toxin attacks a target which is different in eukaryotic and prokaryotic cells there is no need for bacterium to synthesise an immunity protein to protect itself. Many toxins and virulence factors are plasmid coded, but this is not such a general rule as for colicins. In fact the same or closely related toxins may be chromosomal in one strain and plasmid borne in others.

Choleratoxin

Choleratoxin (chromosomal) and the heat-labile enterotoxins of enteropathogenic E. coli (Ent-plasmid coded) are variants of the same toxin. Mechanism essentially the same for both. Effect is loss of water and ions from eukaryotic cells due to massive overproduction of cyclic AMP.

Structure: 3 proteins A1, A2 and B. A1 and A2 are derived from a precursor A protein by proteolytic processing. A1 (23.5 Kd) and A2 (5.5 Kd) are linked by a single disulfide bond. Five B (11.6 Kd) are non-covalently attached to the A subunits to give a ring-like structure, containing one each of A1 and A2 and five B.

Choleratoxin binds to ganglioside GM1 in eukaryotic cell membrane, via B protein. Each B subunit binds to the galactose end of a ganglioside molecule. Free B-protein protects eukaryotic cells against toxin by competing for GM1 binding sites. Structure of GM1 (sphingosine is a base):

GAL - N-Acetyl-Galactosamine - GAL - GLC - Sphingosine - Fatty acid
­
N-Acetyl-Neuraminic Acid

A1 protein activates adenylate cyclase. It must first be released from A2 by reduction of S-S bond after binding of A1-S-S-A2/B5 complex to cytoplasmic membrane of target cell. A1 can then enter the cell. Cyclase activation requires GTP and NAD. Choleratoxin hydrolyses NAD to nicotinamide and ADP-ribose. Originally thought toxic effect of NAD-using toxins (both Cholera and Diptheria - see below) was due to destruction of NAD. However, in vivo these toxins actually transfer the ADP-ribose to another acceptor molecule and this effect is responsible for lethal event. Choleratoxin A1 protein can ADP-ribosylate a variety of acceptors including arginine and its derivatives as well as many proteins. A1 protein can ADP-ribosylate itself and when so modified its activity increases by 50%.

The genuine in vivo lethal target is a specific GTPase involved in regulation of adenylate cyclase. This GTPase is a membrane protein which acts as a regulatory subunit (G-protein) for adenylate cyclase. ADP-ribosylation of an arginine residue on the GTPase results in inactivation of this protein. The cyclase is then freed from the regulation and is permanently activated. Presence of GTP or its nonhydrolysable analog GMP-P-NH-P is necessary for cyclase activation. GTP hydrolysis allows release of regulatory subunit and deactivation. ADP-ribosylation prevents hydrolysis of GTP and hence jams GTPase in GTP-binding state. GTP analogs which cannot be hydrolyzed show similar effects. Cyclic-AMP production rises several hundred fold in cells which are affected. Intestinal cells lose sodium and then water.

Diptheria Toxin

Toxin is encoded by tox genes carried on lysogenic phage rather than plasmids, however principle is similar. Corynebacterium diptheriae strains which lose the phage also lose the ability to produce toxin and, conversely, non-pathogenic strains may be made toxin producers by lysogeny with appropriate phage. The phages b-tox and w-tox have cohesive ends like lambda. They form single, double, or rarely, triple lysogens and the amount of toxin produced corresponds to gene dosage. There are two chromosomal att sites (attB1 and attB2) into which phage integrate at random to give single or double lysogens. Rarely, tandem lysogeny at one of these sites gives an overall triple lysogen. The related phage, g tox, is defective in toxin production.

About 100 nanograms/kg is a lethal dose for most animals, except mice and rats which need 1000 fold more. Toxin is single 63Kd protein that is enzymatically inactive. It is activated by a serine protease which gives A (24Kd) and B (39Kd) fragments, linked by a single disulfide bridge. Fragment A is enzymatically active, B is needed for entry into the target cell.

Fragment A hydrolyses NAD to nicotinamide and ADP-ribose. Lethal event is transfer of the ADP-ribose moiety to translation elongation factor EF-2, and hence inhibition of protein synthesis. EF-2 is a GTP hydrolyzing protein required for moving the peptidyl-tRNA from aminoacyl-site to peptidyl-site on ribosome and shifting the mRNA by one codon in a reaction requiring GTP hydrolysis. ADP-ribosylated EF-2 still binds GTP but cannot hydrolyse it or translocate.

Modification occurs at a diphthamide residue of EF-2. Diphthamide is derived from histidine by post-translational modification and is found in eukaryotes and archebacteria only on EF-2 in part of the amino acid sequence which is highly conserved. Eubacterial factor EF-G does not contain diphthamide (nor do any eubacterial proteins). Archebacterial elongation factor is ADP-ribosylated by diptheria toxin but much less efficiently than eukaryotic EF-2.

Bacteriophage Toxins

Several bacteriophages can ADP-ribosylate bacterial proteins, using NAD, in a manner analogous to cholera and diptheria toxins. Usually several bacterial proteins are modified and the intended target is not known. The purpose may be to inactivate host metabolism by killing key enzymes, or to modify host polymerases so increasing their activity with phage DNA or RNA. Examples include phage T4 which ADP-ribosylates host RNA-polymerase which thereupon loses much of its ability to transcribe E. coli DNA but works well with T4 DNA.