INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Acyclovir (ACV)-resistant herpes simplex viruses (HSV) infections are relatively frequent and can be associated with significant morbidity among immunocompromised patients (14, 15, 17, 37, 38). Resistance to ACV can be the result of point mutations within the viral thymidine kinase (TK) gene, encoding the enzyme responsible for the initial phosphorylation of ACV into ACV-monophosphate or, more rarely, mutations within the viral DNA polymerase (pol) gene (4, 10). The former mechanism is most frequently seen in the clinic (8, 16, 17, 29), probably because TK is not essential for viral replication in most tissues and culture cells (36). However, several reports have demonstrated that TK activity facilitates HSV reactivation from latency in neural ganglia (11, 13, 44, 46).

Changes in the TK gene can result in viruses producing no or partial amounts of TK or with an altered substrate specificity (4, 23). Darby et al. have proposed a preliminary model for the active center of the HSV type 1 (HSV-1) TK enzyme including three distinct regions: an ATP-binding site (amino acids 51 to 63), a nucleoside-binding site (amino acids 168 to 176), and amino acid 336 (12). Indeed, single-point mutations in one or more of these conserved regions have been found in ACV-resistant HSV isolates (16, 20, 25, 30, 41, 42). Furthermore, Sasadeusz et al. have identified mutational hot spots consisting of frameshift mutations within homopolymer nucleotide stretches of G's and C's (41). Recent studies by our group (16) and others (25) have demonstrated that about 50% of the clinical ACV-resistant strains contain an insertion or a deletion of one or two nucleotides in homopolymer runs of G's and C's, whereas the other half presents single-base substitutions in conserved or nonconserved regions of the TK gene.

Characterization of the TK gene from ACV-susceptible isolates frequently reveals the presence of nonsilent mutations outside the active sites of the TK gene, reflecting a certain degree of viral polymorphism (21, 24, 25). At this time, identification of TK mutations conferring resistance to nucleoside analogues is fastidious and requires transfection of the mutated TK gene in a wild-type virus by a rare homologous recombination event followed by selection of the recombinant strain with an antiviral drug. Such drug pressure may frequently lead to generation of new viral mutations not present in vivo, thus limiting the interpretation of the results. It is therefore of major interest to develop alternative methods for rapid identification and efficient evaluation of TK mutations conferring ACV resistance. In this study, we describe a heterologous system using the nonpathogenic protozoan parasite Leishmania tarentolae (normally devoided of TK activity) as a recipient strain for evaluating the role of several viral mutations detected in clinical ACV-susceptible and ACV-resistant HSV strains.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Patients and isolates. HSV clinical isolates from immunocompromised patients (HIV-infected subjects and solid organ transplant recipients) were provided by Sharon Safrin (University of California, San Francisco) (42), the clinical virology laboratory at the University of Minnesota Hospital and Clinics (16), and various hospitals in the Province of Québec, Canada. Upon reception, viruses were grown once on Vero cells, and then stock cultures were stored in aliquots at 80°C.

Antiviral susceptibility assay. Susceptibility to ACV was determined by a plaque reduction assay (PRA) performed on Vero cells (39). Resistance to ACV was defined by a 50% inhibitory drug concentration (IC₅₀) of 8.8 µM (7, 14, 39). Reference laboratory HSV strains KOS (HSV-1) and MS2 (HSV-2) were used as susceptible controls.

Genotypic analysis. Total cellular DNA was extracted from infected Vero cells by the use of the QIAamp Blood Mini Kit (Qiagen, Chatsworth, Calif.). Amplification of the viral TK gene was done in a DNA thermal cycler (Hybaid Omnigene; Interscience, Markham, Ontario, Canada) using ~1 µg of extracted DNA, 1× cloned Pfu DNA polymerase reaction buffer (Stratagene, La Jolla, Calif.), 200 µM each deoxynucleoside triphosphate, 0.2 µM each primer, 2.5 U of PfuTurbo DNA polymerase (Stratagene), and 5% dimethyl sulfoxide. The sequences of the two consensus primers used for both HSV-1 and HSV-2 amplification were 5'-CGTCTAGATGGCGTGAAACTCCCGCACCT-3' (forward) and 5'-ACAAGCTTTCTGTCTTTTTATTGCCGTCAT-3' (reverse), containing the XbaI and HindIII restriction sites (underlined), respectively. Amplification conditions included an initial denaturation step of 5 min at 94°C, followed by 30 cycles of 1 min at 94°C, 1 min at 55°C, and 2 min at 72°C, a final extension step of 10 min at 72°C. PCR products were purified with an extraction kit (QIAquick gel; Qiagen), and then amplified TK genes were directly sequenced using a cycle sequencing kit (Taq DyeDeoxy Terminator; Applied Biosystems, Foster City, Calif.) and a DNA sequencing system (ABI 373A; Applied Biosystems). Results were compared with known TK sequences from reference strain KOS (HSV-1) or 333 (HSV-2) and to sensitive pretherapy isolates from patients when available. All TK mutations were confirmed by double-strand DNA sequencing from two different PCR products.

Cloning and expression of viral TK genes in L. tarentolae. Experiments were conducted using the wild-type L. tarentolae TarII strain described previously (45). Parasites were grown in SDM-79 medium supplemented with 10% fetal bovine serum (Multicell; Wisent Canadian Laboratories, St-Bruno, Québec, Canada) and hemin (5 mg/ml). L. tarentolae TK-recombinant parasites expressing wild-type or mutant TK genes and the neomycin phosphotransferase gene (neo) as a dominant positive selection marker conferring resistance to G418 were generated by transfection of the expression vector pSPNEOTK. The latter plasmid was constructed by first subcloning the SmaI-BamHI NEO expression cassette (34) into the pSP72 vector (Promega, Madison, Wis.) and then the 1.2-kb XbaI-HindIII-digested PCR product containing the TK gene into XbaI-HindIII sites of the pSPNEO vector. Transfection experiments were done as described previously (34), and transfectants were selected and grown in the presence of G418 (40 µg/ml; Geneticin; Life Technologies GIBCO BRL, Gaithersburg, Md.). Expression of the TK and neo genes in vector pSPNEOTK is driven by the -tubulin intergenic region of L. enriettii (22) as shown in Fig. 1. The sequence of the TK gene before and after transfection into Leishmania was confirmed by isolating the expression plasmid with a miniprep kit (Promega) and sequencing the PCR-amplified TK genes.

View larger version (6K): [in this window] [in a new window]   FIG. 1.   Schematic representation of the Leishmania expression vector pSPNEOTK, comprising the HSV TK and neo genes under the control of the intergenic region (IR) of the L. enrietti -tubulin gene (22). Intergenic regions are important for transcript maturation by trans splicing and polyadenylation in the parasite Leishmania (9). The arrow indicates the orientation of transcription for the neo and TK genes.

Susceptibility of Leishmania to GCV. Susceptibility of Leishmania to the nucleoside analogue ganciclovir (GCV; Cytovene, Syntex Laboratories Inc., Palo Alto, Calif.) was determined by using a cell growth assay previously described by Muyombwe et al. (27). Briefly, parasites expressing the HSV TK gene were grown in the presence of various concentrations of GCV (5 to 10,000 µM). After 72 h of incubation, growth of the parasites was assessed by measuring the absorbance of culture medium at 600 nm, and the IC₅₀ was determined.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Susceptibility of HSV clinical isolates to ACV. Eighteen HSV clinical isolates were recovered from 12 patients, including 10 HIV-infected subjects (1-3, 5-10, 12), one solid organ transplant recipient (4), and one subject with undefined underlying illness (11). For the first five patients, a pair of susceptible and resistant isolates was available. For the other seven patients, only ACV-resistant strains were available. Clinical information and phenotypic/genotypic characterization of 13 strains isolated from eight patients (1 to 4, 6, and 10 to 12) have been previously reported (16, 42). Table 1 summarizes susceptibility results to ACV for all clinical isolates. Five HSV isolates recovered from patients either before ACV (n = 2) or after (n = 3) foscarnet therapy were susceptible to ACV (IC₅₀ range, 2 to 4.4 µM), as measured by the PRA method. Thirteen HSV isolates recovered either during or after ACV treatment were resistant to ACV (IC₅₀ range, 9.9 to 966.1 µM). For patients with both ACV-susceptible and ACV-resistant isolates, the mean change in IC₅₀ over time was 22.9 (range, 4.1 to 55.6).

Analysis of HSV TK gene mutations. Compared to pretherapy ACV-susceptible clinical isolates, ACV-resistant isolates from patients 1 to 5 contained at least one distinct mutation within the coding region of the TK gene (Table 1). The first two resistant strains (1b and 2b) had an additional G within a stretch of 7 G's (nucleotides [nt] 433 to 439) compared to their pretherapy counterparts. Isolate 3b had a single-base substitution at nt 1007 which resulted in a Cys-to-Tyr amino acid change at codon 336. This mutation has been previously reported for both laboratory-derived HSV-1 mutants (12, 19, 35) and HSV-2 clinical isolates (41), resulting in a TK-altered or TK-low producer phenotype. Isolate 4b contained a nucleotide substitution (CT) at position 664 that produced a change from Arg to Cys at codon 222. This residue was reported by Balasubramaniam et al. to be part of one of the six conserved regions (site 5, residues 216 to 222) of the TK genes from 12 human and animal herpesviruses (3). Two distinct mutations, Asp55Asn and Arg222His, were present in the proposed ATP-binding site and in conserved region 5 (3), respectively, of isolate 5b but not in the susceptible virus (5a) isolated from the same patient.

Heterologous expression of the HSV TK gene derived from clinical isolates in L. tarentolae and susceptibility to GCV. The role of the mutations found within the HSV TK genes in conferring resistance to nucleoside analogues was evaluated using the nonpathogenic parasite L. tarentolae as a heterologous system. A series of pSPNEOTK vectors carrying TK genes from several ACV-susceptible and ACV-resistant HSV strains were transfected by electroporation into L. tarentolae as described previously (34). Transfection of the pSPNEOTK vector containing TK genes from the ACV-susceptible reference laboratory strains MS2 (HSV-2) and KOS (HSV-1) resulted in parasites expressing resistance to G418 and susceptibility to GCV (IC₅₀s of 55.91 and 17.19 µM, respectively), whereas transfectants without the TK gene were highly resistant to GCV (IC₅₀, >10,000 µM). As shown in Table 2, expression of the TK gene from five ACV-susceptible clinical isolates (1a to 5a) resulted in susceptibility of the parasites to GCV (IC₅₀, <100 µM; range, 11.35 to 99.4 µM). Nonsilent viral mutations from our study that are not associated with ACV resistance are summarized in Table 3. On the other hand, expression of TK genes from 13 ACV-resistant clinical isolates resulted in high-level resistance of the parasites to GCV (IC₅₀, >5,000 µM). More specifically, nucleotide substitutions 3G175-173C, A391C, G527A, and G1007A resulted in Leishmania IC₅₀s exceeding 10,000 µM GCV, whereas the IC₅₀ for the C664T substitution was somewhat lower, 7,993 µM. For two isolates (5b and 8), the simultaneous presence of two TK mutations also resulted in resistance of Leishmania to GCV, although the specific role of each individual mutation could not be assessed. Finally, frameshift mutations (addition or deletion) (strains 1b, 2b, 6a, 6b, 7, and 9) also resulted in high-level resistance of the parasite to GCV. In summary, a >50-fold difference in Leishmania IC₅₀ of GCV (from <100 to >5,000 µM) was found between ACV-susceptible and ACV-resistant HSV isolates.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

ACV-resistant HSV isolates are recovered relatively frequently from immunocompromised subjects receiving this drug for a prolonged period of time (8). Rapid genotypic characterization of drug-resistant mutants directly from clinical biological fluids requires a good knowledge of viral mutations associated with resistance and those associated with viral polymorphism (21, 24, 25). Such distinction is particularly important in the case of point substitutions and when pretherapy isolates are unavailable. Homologous recombination of a mutated TK gene in a wild-type virus following transfection is the method currently in use to confirm the role of specific point mutations. However, homologous recombination is a rare event, and the need to apply a drug pressure to select for such mutants can lead to additional viral mutations. To circumvent this limitation, we developed a heterologous system in which the TK gene from several HSV clinical isolates was expressed in the parasite Leishmania lacking endogenous TK activity.

HSV-1 TK expression has been already used in combination with ACV or GCV as a suicide enzyme in gene therapy for cancer (5, 18, 43), AIDS (6), and Leishmania infection (26, 27). In our model, transfected parasites were evaluated for their susceptibility to GCV, a close analogue of ACV. As described elsewere (27), ACV did not inhibit the growth of Leishmania transfectants expressing TK activity (data not shown), probably due to the different chemical conformation of the molecule precluding its penetration in the parasite. Nevertheless, ACV-resistant HSV strains containing TK mutations have been shown to be cross-resistant to GCV (1, 2, 28). In our study, expression of wild-type HSV TK in Leishmania resulted in GCV IC₅₀s of <100 µM; in contrast, expression of TK genes derived from ACV-resistant HSV strains did not affect the growth of the transfected Leishmania (IC₅₀, >5,000 µM), similar to the situation seen in wild-type parasites. Such a large difference in IC₅₀s using the Leishmania heterologous system can easily clarify any ambiguities between functional mutations and those associated with gene polymorphism (20, 21, 24, 25, 31). In our system, the TK gene is expressed as part of an episomal vector that is present on average at 25 to 50 copies per parasite cell. In such a system, the level of random point mutations that may occur following selection with GCV is expected to be extremely low because it is practically impossible to generate the same type of mutation responsible for a resistance phenotype in all vector copies simultaneously. Moreover, no rearrangements of the TK-transfected genes within the parasite were seen, as confirmed by sequence analysis. Indeed, exogenous DNA transfected into Leishmania, as part of an episomal or an integration vector, is very stable (32, 33).

In the clinic, HSV resistance to nucleoside analogues is usually confirmed by determination of IC₅₀s using the time-consuming and subjective PRA. DNA sequencing of the open reading frame of the TK gene can further confirm the resistance phenotype of an isolate by identifying mutations most likely responsible for ACV resistance. In some cases, the role of the identified mutation in conferring ACV resistance is highly probable. For instance, nucleotide deletions or additions in G and C homopolymer runs have been commonly reported in ACV-resistant isolates (16, 41, 42). These mutations cause a frameshift in the coding sequence of the gene resulting in the formation of a truncated protein as demonstrated by Chatis and Crumpacker (7) and Sasadeusz et al. (41), who performed immunoprecipitation of TK polypeptides and Western blot analysis of virus-infected cell extracts. Expression of such TK mutants in Leishmania resulted in high-level GCV resistance comparable to that of nontransfected parasites, validating the ability of our model to detect nonfunctional TK activity and therefore confirming once again the role of such frameshift mutations in conferring ACV resistance. The role of a specific TK substitution can also be assumed to be associated with ACV resistance in the case where a mutation is present in the resistant isolate but not in a susceptible strain previously recovered from the same patient. However, it is still of interest to ascertain the role of such mutations in isolates containing both TK and DNA pol mutations. Indeed, isolate 6b from this study contained both a TK alteration and a DNA pol mutation between conserved regions I and VII conferring resistance to foscarnet and ACV (42). In this case, we were able to confirm the role of the TK mutation (Leishmania IC₅₀ of GCV of 6,391 µM), although additional experiments are needed to appreciate the overall contribution of the DNA pol mutation in the ACV-resistant phenotype.

It is particularly important to assess the role of viral substitutions in the case where a high degree of gene polymorphism is present and no pretherapy isolates are available. Mutations occurring within the catalytic domains or conserved regions of the enzyme are presumed to be responsible for ACV resistance. However, confirmation of their role in the ACV resistance phenotype may contribute to more precise identification of the importance of specific amino acid residues part of these domains. For example, amino acid changes at residues 59, 176, and 336 in the catalytic sites of HSV-1 and -2 described in this study were shown to induce a GCV-resistant phenotype in Leishmania. On the other hand, some nonsilent TK mutations part of a conserved site can also be associated with gene polymorphism, as illustrated by the GCV-susceptible phenotype in Leishmania associated with change at residue 286 (isolates 3a and 4a) part of the conserved region 6 described by Balasubramaniam et al. (3). Thus, one obvious consequence of our study was to confirm and expand TK gene polymorphisms associated with HSV-1 and -2 (21, 24, 25) (Table 3). Such knowledge could be used to rapidly ascertain the role of specific viral mutations in isolates or directly in clinical samples when some of them have been already associated with gene polymorphism. As an example, the Asn-to-Asp change at position 78 that was found in both ACV-susceptible and -resistant strains from our study has been previously found in an ACV-resistant isolate that also contained a mutation at residue 177 within the nucleoside-binding site (20). Thus, based on our work, the functional role of the TK mutation at codon 78 can be definitively ruled out. We were also able to confirm the role of an amino acid substitution in a noncatalytic and nonconserved region of the TK gene (codon 131) in conferring resistance to ACV. Specific mutations could also be evaluated in our system using site-directed mutagenesis on PCR-amplified TK genes. However, since only the coding region of the TK gene is expressed in the parasite, any mutations occurring outside this region (i.e., within the TK promoter or RNA processing signals) that could affect the expression of the enzyme would not be detected in our system.

One drawback of this system is its inability to determine the levels of resistance conferred by specific TK mutations since all ACV-resistant isolates induced extremely high levels of GCV resistance in Leishmania. Nevertheless, the use of this heterologous system will greatly simplify the identification of mutational hot spots in the TK gene associated with resistance to nucleoside analogues. Moreover, a similar strategy can be used to evaluate the activity of the TK enzyme or its analogue in other herpesviridae (for example varicella-zoster virus [40]) and the impact of TK mutations detected in such viruses.