Single DNA Polymerase Research Articles

Detection of specific RNA targets by multimerization

Detection of specific RNA targets via amplification-mediated techniques is widely used in fundamental studies and medicine due to an essential role of RNA in realization of genetic information and diseases development. Here, we report on an approach for detection of RNA targets based on a particular type of isothermal amplification, namely, reaction of nucleic acid multimerization. The proposed technique requires only a single DNA polymerase possessing reverse transcriptase, DNA-dependent DNA polymerase and strand-displacement activities. Reaction conditions that lead to efficient detection of the target RNAs through multimerization mechanism were determined. The approach was approved using genetic material of SARS-CoV-2 coronavirus as a model viral RNA. Reaction of multimerization allowed to differentiate SARS-CoV-2 RNA-positive samples from SARS-CoV-2 negative samples with high reliability. The proposed technique determines detection of RNA even in samples, which undergone multiple freezing.

Detection of Specific RNA Targets by Multimerization.

Detection of specific RNA targets via amplification-mediated techniques is widely used in fundamental studies and medicine due to essential role of RNA in transfer of genetic information and development of diseases. Here, we report on an approach for detection of RNA targets based on the particular type of isothermal amplification, namely, reaction of nucleic acid multimerization. The proposed technique requires only a single DNA polymerase possessing reverse transcriptase, DNA-dependent DNA polymerase, and strand-displacement activities. Reaction conditions that lead to efficient detection of the target RNAs through multimerization mechanism were determined. The approach was verified by using genetic material of the SARS-CoV-2 coronavirus as a model viral RNA. Reaction of multimerization allowed to differentiate the SARS-CoV-2 RNA-positive samples from the SARS-CoV-2 negative samples with high reliability. The proposed technique allows detection of RNA even in the samples, which were subjected to multiple freezing-thawing cycles.

Ribonucleotide Misincorporation and Reverse Transcriptase Activities of Plasmodium falciparum Apicoplast DNA Polymerase.

Plasmodium falciparumis the most common and harmful causative agent of malaria worldwide. As a member of the phylum Apicomplexa, P. falciparum is characterized by the presence of a unique and essential organelle called the apicoplast. Reminiscent of an algal chloroplast, the apicoplast possesses its own genome, which is maintained by a single apicoplast DNA polymerase (apPol). Ribonucleotides misincorporated into the genome are among the most common lesions encountered by DNA polymerases, and the ability to replicate past these lesions varies widely among characterized enzymes. Here, we have investigated the ribonucleotide (rNTP) misincorporation frequency of apPol and determined its reverse transcriptase (RT) activity across templating ribonucleotides. Pre-steady-state kinetic experiments indicate that apPol does not have an unusually high discrimination between deoxy and ribonucleotides, with frequencies ranging between 104 and 106 depending on the identity of the ribonucleotide. Once incorporated into its template, apPol can replicate across ribonucleotides using its RT activity, but extension of a deoxynucleotide basepaired with the ribonucleotide is slow relative to a canonical basepair. Exonuclease assays indicate that apPol proofreads ribonucleotides an order of magnitude faster than extension, suggesting that most, but not all, misincorporated ribonucleotides will be excised. Although the components have not been identified, ribonucleotide excision repair or other tolerance mechanisms may exist in the P. falciparum apicoplast, and more targeted proteomic efforts will be needed to elucidate them.

Transient State Kinetics of Plasmodium falciparum Apicoplast DNA Polymerase Suggests the Involvement of Accessory Factors for Efficient and Accurate DNA Synthesis.

Plasmodium, the causative agent of malaria, belongs to the phylum Apicomplexa. Most apicomplexans, including Plasmodium, contain an essential nonphotosynthetic plastid called the apicoplast that harbors its own genome that is replicated by a dedicated organellar replisome. This replisome employs a single DNA polymerase (apPol), which is expected to perform both replicative and translesion synthesis. Unlike other replicative polymerases, no processivity factor for apPol has been identified. While preliminary structural and biochemical studies have provided an overall characterization of apPol, the kinetic mechanism of apPol's activity remains unknown. We have used transient state methods to determine the kinetics of replicative and translesion synthesis by apPol and show that apPol has low processivity and efficiency while copying undamaged DNA. Moreover, while apPol can bypass oxidatively damaged lesions, the bypass is error-prone. Taken together, our results raise the following question─how does a polymerase with low processivity, efficiency, and fidelity (for translesion synthesis) faithfully replicate the apicoplast organellar DNA within the hostile environment of the human host? We hypothesize that interactions with putative components of the apicoplast replisome and/or an as-yet-undiscovered processivity factor transform apPol into an efficient and accurate enzyme.

Promising antimalarials targeting apicoplast DNA polymerase from Plasmodium falciparum

Malaria is caused by the parasite Plasmodium falciparum, which contains an essential non-photosynthetic plastid called the apicoplast. A single DNA polymerase, apPOL, is targeted to the apicoplast, where it replicates and repairs the genome. apPOL has no direct orthologs in mammals and is considered a promising drug target for the treatment and/or prevention of malaria. We previously reported screening the Malaria Box to identify MMV666123 as an inhibitor of apPOL. Herein we extend our studies and report structure-activity relationships for MMV666123 and identify key structural motifs necessary for inhibition. Although attempts to crystallize apPOL with the inhibitor were not fruitful, kinetic analysis and crystal structure determinations of WT and mutant apo-enzymes, facilitated model building and provided insights into the putative inhibitor binding site. Our results validate apPOL as an antimalarial target and provide an avenue for the design of high potency, specific inhibitors of apPOL and other A-family DNA polymerases.

Host translesion polymerases are required for viral genome integrity

Human cells encode up to 15 DNA polymerases with specialized functions in chromosomal DNA synthesis and damage repair. In contrast, complex DNA viruses, such as those of the herpesviridae family, encode a single B-family DNA polymerase. This disparity raises the possibility that DNA viruses may rely on host polymerases for synthesis through complex DNA geometries. We tested the importance of error-prone Y-family polymerases involved in translesion synthesis (TLS) to human cytomegalovirus (HCMV) infection. We find most Y-family polymerases involved in the nucleotide insertion and bypass of lesions restrict HCMV genome synthesis and replication. In contrast, other TLS polymerases, such as the polymerase ζ complex, which extends past lesions, was required for optimal genome synthesis and replication. Depletion of either the polζ complex or the suite of insertion polymerases demonstrate that TLS polymerases suppress the frequency of viral genome rearrangements, particularly at GC-rich sites and repeat sequences. Moreover, while distinct from HCMV, replication of the related herpes simplex virus type 1 is impacted by host TLS polymerases, suggesting a broader requirement for host polymerases for DNA virus replication. These findings reveal an unexpected role for host DNA polymerases in ensuring viral genome stability.

Phage Origin of Mitochondrion-Localized Family A DNA Polymerases in Kinetoplastids and Diplonemids.

Mitochondria retain their own genomes as other bacterial endosymbiont-derived organelles. Nevertheless, no protein for DNA replication and repair is encoded in any mitochondrial genomes (mtDNAs) assessed to date, suggesting that the nucleus primarily governs the maintenance of mtDNA. As the proteins of diverse evolutionary origins occupy a large proportion of the current mitochondrial proteomes, we anticipate finding the same evolutionary trend in the nucleus-encoded machinery for mtDNA maintenance. Indeed, none of the DNA polymerases (DNAPs) in the mitochondrial endosymbiont, a putative α-proteobacterium, seemingly had been inherited by their descendants (mitochondria), as none of the known types of mitochondrion-localized DNAP showed a specific affinity to the α-proteobacterial DNAPs. Nevertheless, we currently have no concrete idea of how and when the known types of mitochondrion-localized DNAPs emerged. We here explored the origins of mitochondrion-localized DNAPs after the improvement of the samplings of DNAPs from bacteria and phages/viruses. Past studies have revealed that a set of mitochondrion-localized DNAPs in kinetoplastids and diplonemids, namely PolIB, PolIC, PolID, PolI-Perk1/2, and PolI-dipl (henceforth designated collectively as “PolIBCD+”) have emerged from a single DNAP. In this study, we recovered an intimate connection between PolIBCD+ and the DNAPs found in a particular group of phages. Thus, the common ancestor of kinetoplastids and diplonemids most likely converted a laterally acquired phage DNAP into a mitochondrion-localized DNAP that was ancestral to PolIBCD+. The phage origin of PolIBCD+ hints at a potentially large contribution of proteins acquired via nonvertical processes to the machinery for mtDNA maintenance in kinetoplastids and diplonemids.

Tracking DNA Synthesis with Single-Molecule Strand Displacement.

We have previously shown that double-stranded DNA labeled with a periodic series of fluorescent dyes can be used to track a single helicase. Here we demonstrate how this technique can be adapted to follow processive DNA synthesis. By monitoring strand displacement, we track the motion of a single ϕ29 DNA polymerase without labeling or altering the enzyme or the template strand, and without applying any force. We observe a wide range of speeds, with the highest exceeding by several times those observed in earlier in vitro single-molecule experiments. Because this method enables repeated observations of the same polymerase traversing identical segments of DNA, it should prove useful for determining the effects of sequence on DNA replication and transcription. In addition, future measurements of this type may allow us to examine in detail the interactions of individual DNA polymerases with other components of the replisome.

Influence of DNA Lesions on Polymerase-Mediated DNA Replication at Single-Molecule Resolution.

Faithful replication of DNA is a critical aspect in maintaining genome integrity. DNA polymerases are responsible for replicating DNA, and high-fidelity polymerases do this rapidly and at low error rates. Upon exposure to exogenous or endogenous substances, DNA can become damaged and this can alter the speed and fidelity of a DNA polymerase. In this instance, DNA polymerases are confronted with an obstacle that can result in genomic instability during replication, for example, by nucleotide misinsertion or replication fork collapse. It is important to know how DNA polymerases respond to damaged DNA substrates to understand the mechanism of mutagenesis and chemical carcinogenesis. Single-molecule techniques have helped to improve our current understanding of DNA polymerase-mediated DNA replication, as they enable the dissection of mechanistic details that can otherwise be lost in ensemble-averaged experiments. These techniques have also been used to gain a deeper understanding of how single DNA polymerases behave at the site of the damage in a DNA substrate. In this review, we evaluate single-molecule studies that have examined the interaction between DNA polymerases and damaged sites on a DNA template.

Nuclear and mitochondrial DNA analysis reveals that hybridization between Fasciola hepatica and Fasciola gigantica occurred in China.

The well-known pathogens of fasciolosis, Fasciola hepatica (Fh) and Fasciola Gigantica (Fg), possess abundant mature sperms in their seminal vesicles, and thus, they reproduce bisexually. On the other hand, aspermic Fasciola flukes reported from Asian countries, which have no sperm in their seminal vesicles, probably reproduce parthenogenetically. The aim of this study was to reveal the origin of aspermic Fasciola flukes. The nuclear single copy markers, phosphoenolpyruvate carboxykinase and DNA polymerase delta, were employed for analysis of Fasciola species from China. The hybrid origin of aspermic Fasciola flukes was strongly suggested by the presence of the Fh/Fg type, which includes DNA fragments of both F. hepatica and F. gigantica. China can be regarded as the cradle of the interspecific hybridization because F. hepatica and F. gigantica were detected in the northern and southern parts of China, respectively, and hybrids flukes were distributed between the habitats of the two species. The Chinese origin was supported by the fact that a larger number of mitochondrial NADH dehydrogenase subunit 1 (nad1) haplotypes was detected in Chinese aspermic Fasciola populations than in aspermic populations from the neighbouring countries. Hereafter, 'aspermic' Fasciola flukes should be termed as 'hybrid' Fasciola flukes.

Mechanisms of mutagenesis: DNA replication in the presence of DNA damage.

Environmental mutagens cause DNA damage that disturbs replication and produces mutations, leading to cancer and other diseases. We discuss mechanisms of mutagenesis resulting from DNA damage, from the level of DNA replication by a single polymerase to the complex DNA replisome of some typical model organisms (including bacteriophage T7, T4, Sulfolobus solfataricus, Escherichia coli, yeast and human). For a single DNA polymerase, DNA damage can affect replication in three major ways: reducing replication fidelity, causing frameshift mutations, and blocking replication. For the DNA replisome, protein interactions and the functions of accessory proteins can yield rather different results even with a single DNA polymerase. The mechanism of mutation during replication performed by the DNA replisome is a long-standing question. Using new methods and techniques, the replisomes of certain organisms and human cell extracts can now be investigated with regard to the bypass of DNA damage. In this review, we consider the molecular mechanism of mutagenesis resulting from DNA damage in replication at the levels of single DNA polymerases and complex DNA replisomes, including translesion DNA synthesis.

A Jump-from-Cavity Pyrophosphate Ion Release Assisted by a Key Lysine Residue in T7 RNA Polymerase Transcription Elongation.

Pyrophosphate ion (PPi) release during transcription elongation is a signature step in each nucleotide addition cycle. The kinetics and energetics of the process as well as how it proceeds with substantial conformational changes of the polymerase complex determine the mechano-chemical coupling mechanism of the transcription elongation. Here we investigated detailed dynamics of the PPi release process in a single-subunit RNA polymerase (RNAP) from bacteriophage T7, implementing all-atom molecular dynamics (MD) simulations. We obtained a jump-from-cavity kinetic model of the PPi release utilizing extensive nanosecond MD simulations. We found that the PPi release in T7 RNAP is initiated by the PPi dissociation from two catalytic aspartic acids, followed by a comparatively slow jump-from-cavity activation process. Combining with a number of microsecond long MD simulations, we also found that the activation process is hindered by charged residue associations as well as by local steric and hydrogen bond interactions. On the other hand, the activation is greatly assisted by a highly flexible lysine residue Lys472 that swings its side chain to pull PPi out. The mechanism can apply in general to single subunit RNA and DNA polymerases with similar molecular structures and conserved key residues. Remarkably, the flexible lysine or arginine residue appears to be a universal module that assists the PPi release even in multi-subunit RNAPs with charge facilitated hopping mechanisms. We also noticed that the PPi release is not tightly coupled to opening motions of an O-helix on the fingers domain of T7 RNAP according to the microsecond MD simulations. Our study thus supports the Brownian ratchet scenario of the mechano-chemical coupling in the transcription elongation of the single-subunit polymerase.

Yeast model analysis of novel polymerase gamma variants found in patients with autosomal recessive mitochondrial disease.

Replication of the mitochondrial genome depends on the single DNA polymerase (pol gamma). Mutations in the POLG gene, encoding the catalytic subunit of the human polymerase gamma, have been linked to a wide variety of mitochondrial disorders that show remarkable heterogeneity, with more than 200 sequence variants, often very rare, found in patients. The pathogenicity and dominance status of many such mutations remain, however, unclear. Remarkable structural and functional conservation of human POLG and its S. cerevisiae ortholog (Mip1p) led to the development of many successful yeast models, enabling to study the phenotype of putative pathogenic mutations. In a group of patients with suspicion of mitochondrial pathology, we identified five novel POLG sequence variants, four of which (p.Arg869Ter, p.Gln968Glu, p.Thr1053Argfs*6, and p.Val1106Ala), together with one previously known but uncharacterised variant (p.Arg309Cys), were amenable to modelling in yeast. Familial analysis indicated causal relationship of these variants with disease, consistent with autosomal recessive inheritance. To investigate the effect of these sequence changes on mtDNA replication, we obtained the corresponding yeast mip1 alleles (Arg265Cys, Arg672Ter, Arg770Glu, Thr809Ter, and Val863Ala, respectively) and tested their effect on mitochondrial genome stability and replication fidelity. For three of them (Arg265Cys, Arg672Ter, and Thr809Ter), we observed a strong, partially dominant phenotype of a complete loss of functional mtDNA, whereas the remaining two led to partial mtDNA depletion and significant increase in point mutation frequencies. These results show good correlation with the severity of symptoms observed in patients and allow to establish these variants as pathogenic mutations.Electronic supplementary materialThe online version of this article (doi:10.1007/s00439-015-1578-x) contains supplementary material, which is available to authorized users.

First-passage problems in DNA replication: effects of template tension on stepping and exonuclease activities of a DNA polymerase motor

A DNA polymerase (DNAP) replicates a template DNA strand. It also exploits the template as the track for its own motor-like mechanical movement. In the polymerase mode it elongates the nascent DNA by one nucleotide in each step. However, whenever it commits an error by misincorporating an incorrect nucleotide, it can switch to an exonuclease mode. In the latter mode it excises the wrong nucleotide before switching back to its polymerase mode. We develop a stochastic kinetic model of DNA replication that mimics an in vitro experiment where single-stranded DNA, subjected to a mechanical tension F, is converted to double-stranded DNA by a single DNAP. The F-dependence of the average rate of replication, which depends on the rates of both polymerase and exonuclease activities of the DNAP, is in good qualitative agreement with the corresponding experimental results. We introduce nine novel distinct conditionaldwell times of a DNAP. Using the method of first-passage times, we also derive the exact analytical expressions for the probability distributions of these conditional dwell times. The predicted F-dependences of these distributions are, in principle, accessible to single-molecule experiments.

207 Nanopore immobilization of DNA polymerase enhances single-molecule sequencing

A zero-mode waveguide (ZMW) is a nanoscale optical waveguide driven at a frequency below its cut-off. In this mode, the electric field, instead of traveling down the axis of the conducting cavity, decays exponentially. By fabricating waveguides with sub-wavelength diameters and illuminating them with laser light, the electric field in the waveguide is confined enough to enable single-molecule optical detection at micromolar concentration [1]. Immobilizing single DNA polymerases in ZMWs and using special phosphate-fluorescently labeled dNTPs form the basis for single-molecule real-time DNA sequencing, one of the most promising next-generation sequencing platforms [2]. In this method, the polymerase replicates the sample DNA, and as it incorporates new bases into the product strand, the labeled dNTPs emit a burst of light before the phosphate is cleaved off. The sequence of colors corresponds to the DNA sequence (see Figure 1 below from Eid et al., 2009). Because the ZMW aperture’s diameter is sub-diffraction-limit, it is impossible to optically distinguish one polymerase in a ZMW from two. Having only one polymerase in each waveguide is critical to sequencing accuracy. In its present state, experimenters use diffusion to fill ZMWs with polymerases, resulting in a Poisson distribution for filling ZMWs, and consequently a theoretical limit of 36.8% of ZMWs having only one polymerase [2]. We achieve full polymerase occupancy of ZMWs by fabricating the structures on an ultrathin silicon nitride membrane and drilling a nanopore at the base of each waveguide with an ion beam. A short DNA fragment with biotin on either end is conjugated to a streptavidin and then drawn into the nanopore with a voltage bias. There is then a free biotin at the base of the ZMW. A polymerase–streptavidin complex can diffuse into the ZMW and bind to the exposed biotin. Because the nanopore is too small to fit more than one molecule, only one ZMW will bind to a biotin in the nanopore. Upon flushing the ZMW chamber, the biotin-bound polymerase will remain trapped in the pore, and only a single polymerase will remain at the base of each waveguide.

Human mitochondrial DNA replication.

Elucidation of the process of DNA replication in mitochondria is in its infancy. For many years, maintenance of the mitochondrial genome was regarded as greatly simplified compared to the nucleus. Mammalian mitochondria were reported to lack all DNA repair systems, to eschew DNA recombination, and to possess but a single DNA polymerase, polymerase γ. Polγ was said to replicate mitochondrial DNA exclusively via one mechanism, involving only two priming events and a handful of proteins. In this "strand-displacement model," leading strand DNA synthesis begins at a specific site and advances approximately two-thirds of the way around the molecule before DNA synthesis is initiated on the "lagging" strand. Although the displaced strand was long-held to be coated with protein, RNA has more recently been proposed in its place. Furthermore, mitochondrial DNA molecules with all the features of products of conventional bidirectional replication have been documented, suggesting that the process and regulation of replication in mitochondria is complex, as befits a genome that is a core factor in human health and longevity.

The Cost of Being Right During Replication

DNA replication needs to be both faithful and processive. DNA polymerases replicate DNA, using a template strand to add the complementary nucleotide to the synthesizing primer (polymerization). Family A DNA polymerases, like T7 DNA polymerase (DNAp), boost their fidelity by proofreading the primer. In the event of a mismatch, the erroneous nucleotide is excised at the exonuclease active site (proofreading). Proper balance between polymerization and proofreading is essential to achieve faithful but productive replication.Here we report on a single-molecule study of DNAp using optical tweezers. In our setup we can directly measure the individual rates of replicating, proofreading and pausing single DNA polymerases at various tensions. With a spatial and temporal resolution of 25 bp and 0.2 sec, we can study DNAp in far more detail than previous single-molecule studies.We find that the rate of polymerization decreases with tension on the DNA but, notably, doesn’t stall at high tension. Next, the rate of processive proofreading is constant over all measured tensions which, surprisingly, include tensions of only a few pN. This means that the proofreading state of DNAp is highly accessible, independent of perturbations of the primer-template structure, and that during proofreading tens of nucleotides at a time are removed.Our observations indicate that the current kinetic model for DNAp is not sufficient. The length of the pauses in between events are distributed as a double exponential, arguing against a single inactive state. Therefore, we introduce an updated model that contains an extra state that might play a role in balancing polymerization and proofreading. The conclusion seems that the cost of being right (i.e. error-free replication) outweighs the price a cell pays (i.e. the regular removal of correct base pairs) for picking up errors during replication.

Tracking Processive DNA Synthesis with Single-Molecule Fluorescence

We have recently shown that double-stranded DNA labeled with a periodic series of fluorescent dyes can be used to track a single helicase. Here we demonstrate that this technique can be adapted to follow processive DNA synthesis. By monitoring strand displacement, we measure the motion of a single phi29 DNA polymerase without labeling or altering the enzyme or the template strand, and without applying any force. We observe a wide range of speeds, with the highest exceeding by several times that observed in other single-molecule experiments. Because this method enables repeated observations of the same polymerase traversing identical segments of DNA, it should prove useful for determining the effects of sequence on DNA replication and transcription.

Human replicative DNA polymerase δ can bypass T‐T (6‐4) ultraviolet photoproducts on template strands

DNA polymerase δ (Polδ) carries out DNA replication with extremely high accuracy. This great fidelity primarily depends on the efficient exclusion of incorrect base pairs from the active site of the polymerase domain. In addition, the 3'-5' exonuclease activity of Polδ further enhances its accuracy by eliminating misincorporated nucleotides. It is believed that these enzymatic properties also inhibit Polδ from inserting nucleotides opposite damaged templates. To test this widely accepted idea, we examined in vitro DNA synthesis by human Polδ enzymes proficient and deficient in the exonuclease activity. We chose the UV-induced lesions cyclobutyl pyrimidine dimer (CPD) and 6-4 pyrimidone photoproduct (6-4 PP) as damaged templates. 6-4 PP represents the most formidable challenge to DNA replication, and no single eukaryotic DNA polymerase has been shown to bypass 6-4 PP in vitro. Unexpectedly, we found that Polδ can perform DNA synthesis across both 6-4 PP and CPD even with a physiological concentration of deoxyribonucleotide triphosphates (dNTPs). DNA synthesis across 6-4 PP was often accompanied by a nucleotide deletion and was highly mutagenic. This unexpected enzymatic property of Polδ in the bypass of UV photoproducts challenges the received notion that the accuracy of Polδ prevents bypassing damaged templates.

Opposite effect of two cytomegalovirus DNA polymerase mutations on replicative capacity and polymerase activity

Human cytomegalovirus (HCMV) infections cause significant morbidity in immunocompromised hosts. The aim of this study was to characterize the role of two HCMV DNA polymerase mutations (K805Q and T821I) found in a ganciclovir- and foscarnet-resistant clinical isolate from an AIDS patient. The effects of single and dual DNA polymerase mutations on virus susceptibility and replicative capacity, as well as on enzymatic activity, were studied using recombinant viruses generated from overlapping cosmids and DNA polymerase enzymes expressed in rabbit reticulocyte lysates. Recombinant viruses containing mutations K805Q, T821I and K805Q+T821I had 0.8-fold, 5.3-fold and 4.8-fold increases in ganciclovir 50% inhibitory concentration (IC(50)) values and 0.3-fold, 23.3-fold and 15.6-fold increases in foscarnet IC(50) values, respectively, compared with those of the wild-type virus. The recombinant virus T821I had impaired replication in fibroblastic cells on day 2 post-infection with a decrease in viral titres of 3.5-fold, 4.3-fold and 2.6-fold compared to the recombinant wild-type, K805Q and K805Q+T821I viruses, respectively. Enzymatic studies of wild-type and mutant DNA polymerase enzymes in presence of foscarnet resulted in IC(50) values that were similar to those of the recombinant viruses. Steady-state kinetic constants K(m) and V(max) derived from Michaelis-Menten equations showed that the activity of the mutant T821I enzyme was diminished compared with those of wild-type, K805Q and K805Q+T821I mutant enzymes. Thermodynamic stability of the two single mutant enzymes was opposed as shown by computer-assisted three-dimensional modelling studies. The HCMV DNA polymerase mutation K805Q improved the fitness of the T821I mutation associated with high levels of resistance to foscarnet.