Utilization of Microsatellite Polymorphism for Differentiating Herpes Simplex Virus Type 1 Strains

Abstract

The herpes simplex virus type 1 (HSV-1) genome is a linear double-stranded DNA of 152 kpb. It is divided into long and short regions of unique sequences termed U_L and U_S, respectively, and these are flanked by regions of inverted internal and terminal repeats. Microsatellites are short tandem repeats of 1- to 6-nucleotide motifs; they are often highly variable and polymorphic within the genome, which raises the question of whether they may be used as molecular markers for the precise differentiation of HSV-1 strains. In this study, 79 different microsatellites (mono-, di-, and trinucleotide repeats) in the HSV-1 complete genome were identified by in silico analysis. Among those microsatellites, 45 were found to be distributed in intergenic or noncoding inverted repeat regions, while 34 were in open reading frames. Length polymorphism analysis of the PCR products was used to investigate a set of 12 distinct HSV-1 strains and allowed the identification of 23 polymorphic and 6 monomorphic microsatellites, including two polymorphic trinucleotide repeats (CGT and GGA) within the UL46 and US4 genes, respectively. A multiplex PCR method that amplified 10 polymorphic microsatellites was then developed for the rapid and accurate genetic characterization of HSV-1 strains. Each HSV-1 strain was characterized by its own microsatellite haplotype, which proved to be stable over time in cell culture. This relevant innovative tool was successfully applied both to confirm the close relationship between sequential HSV-1 isolates collected from patients with multiple recurrent infections and to investigate putative nosocomial infections.

Herpes simplex virus type 1 (HSV-1) is a member of the subfamily Alphaherpesvirinae. The seroprevalence of HSV-1 infection increases progressively from childhood and is inversely proportional to an individual's socioeconomic background (35). Primary HSV-1 infections in children are typically asymptomatic but can give rise to herpetic gingivostomatitis. After primary infection of the orofacial region, HSV-1 is transported in a retrograde manner to the nuclei of the trigeminal sensory neurons through their axons, which innervate the infected area. HSV-1 then establishes a life-long latent infection in the nuclei of sensory neurons, where the genome lies in a nonreplicating chromatin-associated state. Recurrent HSV-1 lesions occur following the reactivation of latent HSV-1, axonal transport of the reactivated virus, and HSV-1 replication on the skin and mucous membranes. Recurrent infections typically give rise to herpes labialis or may be responsible for more severe clinical manifestations, including keratitis, meningoencephalitis, bronchopneumonitis (22), chronic or disseminated infections in immunosuppressed patients, or eczema herpeticum. Eczema herpeticum, or Kaposi-Juliusberg disease, is an uncommon herpes simplex virus superinfection that occurs in patients with atopic dermatitis. Additionally, HSV-1 accounts for about half of the new cases of genital herpes in developed countries (14).

The HSV-1 genome is a linear double-strand DNA molecule of 152,261 bp (strain 17, GenBank accession number X14112) that encodes at least 77 genes (24, 28). It is composed of two covalently linked regions of unique sequences, termed the unique long (U_L) and unique short (U_S) sequences, which are bracketed by inverted repeats. Moreover, HSV-1 U_S and U_L undergo inversions resulting from recombination events that yield four genomic isomers in equimolar accounts (17).

The differentiation of unrelated HSV-1 strains has been used to trace the spread of infection in cases of nosocomial outbreaks (1, 6, 16); pathogenesis investigations (31, 32); and epidemiological studies, which also suggest that HSV-1 strains among the same geographic or ethnic groups are evolutionarily more closely related (5, 37, 39). HSV-1 strains differ slightly in their DNA sequences, which are used as molecular markers of genomic polymorphism. The method generally used to differentiate HSV-1 strains is restriction fragment length polymorphism (RFLP) analysis of total HSV-1 DNA (4, 7, 26, 36) or RFLP-PCR-based analysis (23, 49, 53). Determination of the nucleotide sequences of PCR-amplified DNA fragments encompassing reiterated regions (33) is another approach currently used to characterize the genomes of HSV-1 strains. Differences in the sizes of specific fragments among strains are detected in a PCR-based assay and are attributed to a variation in copy numbers or in the nucleotide sequences of these hypervariable reiterations, which have been identified within the ICP34.5 gene (8), the “a” sequence (10), the US10 and US11 genes, the introns of the US1 and US12 genes (23, 47, 48), and the glycoprotein G (gG) and gI genes (27).

Microsatellites are short tandem repeats of 1 to 6 bp and are the most mutable DNA sequences found in eukaryote, prokaryote, and some virus genomes, particularly in intergenic regions and introns (11, 43). While they are hypermutable, microsatellites have been used as molecular markers in numerous ecology, evolutionary genetics, and genome mapping studies. Moreover, the instability of these microsatellites, which are mainly homopolymer nucleotide repeats, may improve frameshift mutations that provide phenotypic changes and an adaptive advantage in the evolution of some mutated viral strains (12). The identification and genetic mapping of highly polymorphic microsatellite loci in human cytomegalovirus (HCMV) have been reported previously (9). Therefore, microsatellite polymorphism analysis proved to be a rapid and accurate tool for studying the epidemiology of HCMV infections (29, 51). To our knowledge, this strategy has never been applied to HSV-1 infections.

The aim of our study was to establish the map and characterize the short microsatellites within the full-length HSV-1 genome and to confirm the utility of microsatellites as precise molecular markers for the differentiation of HSV-1 strains.

MATERIALS AND METHODS

Cells and virus strains.

Vero cells (African green monkey kidney cells) and human embryonic lung fibroblasts (MRC5 cells) were propagated in minimum essential medium (GIBCO, Cergy, France) supplemented with 3% fetal bovine serum (GIBCO), amikacin (20 μg/ml; Bristol-Myers Squibb, Rueil-Malmaison, France), vancomycin (20 μg/ml; Lilly, Suresnes, France), glutamine (1%; Eurobio, Courtaboeuf, France), and a nonessential amino acid solution (1%; Sigma, Saint-Quentin Fallavier, France), referred to hereafter as the culture medium.

A total of 21 isolates and one laboratory strain (strain KOS) were used in this study. The clinical isolates were obtained on Vero cells from oral, bronchoalveolar lavage, or cutaneous specimens from unrelated patients submitted for the diagnosis of clinical herpes reactivations. Determination of the HSV type of each strain was performed by means of an immunofluorescence assay with specific monoclonal antibodies, as described previously (44).

Microsatellite identification.

The MsatFinder program was used to search the HSV-1 whole-genome sequence data (strain 17, GenBank accession number X14112) for microsatellites (45). The output results were filtered to include only short microsatellites with at least four trinucleotide, five dinucleotide, or nine mononucleotide repeats units, as previously described by Davis et al. (9).

Microsatellite characterization by sequencing.

DNA was extracted from HSV-1-infected cells exhibiting a full cytopathic effect by using a QIamp DNA blood kit (Qiagen, Courtaboeuf, France), according to the manufacturer's instructions. We analyzed 29 microsatellites using 25 PCR primer pairs (Table 1). PCRs containing a reaction mixture with 1× buffer (Invitrogen), 2 mM MgCl₂, 0.2 mM each deoxynucleoside triphosphate, 200 nM each forward and reverse primers, and 0.2 U of proofreading Taq polymerase (Platinum, Invitrogen) were performed in a Mastercycler ep thermal cycler (Eppendorf, France). The reaction conditions were 1 cycle of denaturation (2 min at 96°C), 30 cycles (30 s at 96°C, 60 s at 55°C, 60 s at 68°C), and a final elongation step (10 min at 68°C). Negative controls were included in each run, and the specificity of the PCR was evaluated by including in each PCR run DNA extracted from noninfected Vero and MRC5 cells. Sequencing was carried out on a 3100 ABI sequencer, and all reported sequences were determined twice in each direction. Before sequencing of the PCR products, they were purified by using a Microcon kit (Millipore).

TABLE 1.

Characteristics of amplicons containing HSV-1 microsatellite sequences

Microsatellite namea	Primer sequenceb	Expected sizec (bp)	Target repeat	Positiond	Locuse	No. of variantsf
HSV1-M5	F, ATACACATGGCCCCTTTGAC	477	(G)9	2997-3005	RL2	7
	R, CAGGGAAACACCCAGACATC
HSV1-M10	F, CTATAGGGCAACAAAGGATGG	174	(G)11	6181-6191	TRL	6
	R, CTGTGATTCTCTGGCTGCAC
HSV1-M11	F, TCCTCCGTATCCTCGCTTTA	182	(C)10	6561-6570	TRL	7
	R, CAATGACCCACGTACTCCAA
HSV1-M14	F, ACCACGTGGGTAGGTGATGT	229	(G)10	7794-7803	TRL	5
	R, AGAAAGGCCCAGAGTCATTG
HSV1-20/21	F, TTCGGGTCGTTATTCAGTCC	396	(C)10	11668-11677	Inter-UL3-UL4	8 (2)
	R, ACCTGCTGGTAGAGGTGCTG		(G)11	11772-11782
HSV1-M22	F, GTAATGGCGCTGACGCTCT	300	(G)10	20563-20572	Inter-UL8-UL9	6
	R, CGGTGAACTTTAGCACCCTA
HSV1-M23	F, CGCCGTTGGTAGCTGTTT	410	(C)13	24717-24729	Inter-UL10-UL11	7
	R, GGAGGTCGTCTCGCTGAC
HSV1-M24	F, GACTCCATTTCCGAGACGAC	398	(TG)5	27053-27062	UL13	1
	R, GGTTTCCCGGTTACCAGTATT
HSV1-M25	F, TATTCCGGAAAACGGAACG	369	(G)12	34869-34880	Inter-UL15-UL18	5
	R, GGCATCCAGGGAGAAACC
HSV1-M26	F, CAAGGTCCACCGAAAGACAC	378	(C)11	40797-40807	Inter-UL19-UL20	5
	R, TTTATCGCCACCAACCAAAT
HSV1-M28	F, GAACTGGGCGTAGTTCGTG	386	(GC)5	45585-45594	UL22	1
	R, TGTTGAGATCCCCAGGTAGG
HSV1-M30	F, GGAGCTGGGTTAGCGTTT	143	(T)11	50769-50779	Inter-UL25-UL26	7
	R, CTCCTCCATCCGGTCTCC
HSV1-M31/32	F, GCCGATTTGTTCGTCTCTCA	452	(G)12	52880-52891	Inter-UL26.5-UL27	9
	R, ACACCCAAGTTCCCAACAAA		(CA)5	52991-53000
HSV1-M33	F, CACGTTCTCGACGCTGTAAT	384	(GC)6	56495-56506	UL28	1
	R, CTGACGAAGCGGTCGTTG
HSV1-M34	F, GGGACCTTGATGGTGGTTG	294	(G)9	62142-62150	Inter-UL29-UL30	6
	R, GCACCTGACCGTAAGCATCT
HSV1-M36	F, GGCACACCAGACACCAAAG	479	(GC)5	72495-72504	UL36	1
	R, CTCGCGTCTGACGACTCC
HSV1-M38	F, TTCCCCTGATTAGGCTACGA	223	(C)11	80640-80650	Inter-UL36-UL37	7
	R, GTCGAGGACGACGAGATGAT
HSV1-M39	F, GCCATTACCTCTCGGATGG	243	(CG)5	81328-81337	UL37	1 (2)
	R, GAAGCTGCTGTCGGGAGAC
HSV1-M41	F, CACGCAGACGTTTTCGTAGA	369	(CGT)6	99188-99205	UL46	2
	R, GCCACGTATTACACGCACAT
HSV1-M42	F, CTGCGGTTCAGCTCCTGT	248	(ACG)4	111131-111142	UL52	1
	R, GCACAGGTCAAAGTAGGCAAC
HSV1-M48	F, ACTCACAAGCGGCAGAGG	248	(C)10	118568-118577	IRL	5
	R, ACCACGTGGGTAGGTGATGT
HSV1-M49/50	F, ATGAACACTCGGGGTTACCG	297	(T)9	119301-119309	IRL	5
	R, GCAGGTGTCTAACCTACCTGGA		(C)10	119356-119365
HSV1-M51/52	F, CAATGACCCACGTACTCCAA	581	(G)10	119801-119810	IRL	8 (2)
	R, CTATAGGGCAACAAAGGATGG		(C)11	120180-120190
HSV1-M69	F, GTTCCCACCAACGTCTCCT	246	(GGA)4	136977-136988	US4	2
	R, GGACTGGGGTAGGGTGTCA
HSV1-M70	F, GCATGTCTGCTTGAGCTCCT	242	(C)15	138013-138027	Inter-US5-US6	5
	R, CTGGTTAGCCCGGTGTTGT

^aThe 10 microsatellites in boldface are those that were selected for testing by multiplex PCR.

^bF, forward; R, reverse.

^cExpected length of the PCR fragment according to the annotation of the reference HSV-1 complete sequence (strain 17, GenBank accession number X14112).

^dAccording the annotation of the same reference sequence.

^eA gene name in italics indicates that the target repeat was located within an open reading frame, according to the annotation of the same reference sequence.

^fNumber of variant alleles among the panel of the first 12 HSV-1 strains studied, as determined by both length analysis and nucleotide sequencing. M21, M39, and M51 were considered impure microsatellites, since some strains presented both length and sequence polymorphisms. The numbers in parentheses indicate the number of sequence variants.

Microsatellite characterization by length polymorphism analysis.

Each PCR was performed as described above by using a forward primer that was labeled at the 5′ end with a fluorescent phosphoramidite dye (6-carboxyfluorescein). One microliter of each PCR product was diluted with distilled water (1:300) and prepared by adding 0.5 μl of an internal size standard (GeneScan-500 ROX or GeneScan-600 ROX; Applied Biosystems) and highly purified deionized formamide (Applied Biosystems) for a final volume of 16 μl. One previously sized control sample was amplified and run on each gel to ensure that the amplicons were sized correctly. All products were analyzed on 3730 ABI sequencer (Applied Biosystems), according to the manufacturer's instructions. Fragment sizing was performed by using GeneMapper software (version 4.0; Applied Biosystems).

Multiplex PCR.

Ten primer pairs that were able to be multiplexed into the same PCR run and that were able to produce amplicons simultaneously analyzable within a single gel lane were then selected (Table 1). Multiplex PCRs were performed in a Mastercycler ep thermal cycler (Eppendorf) with a PCR mixture containing 0.7× buffer (Invitrogen); 4 mM MgCl₂; 0.2 mM each deoxynucleoside triphosphate; 240 nM primer M69; 400 nM primers M5 and M22; 500 nM primer M23; 600 nM primers M30, M10, M41, M26, and M31/32; 1 μM primer M14; and 0.3 U of proofreading Taq polymerase (Platinum; Invitrogen). The reaction conditions were 1 cycle of denaturation (2 min at 96°C), 30 cycles (30 s at 96°C, 60 s at 65°C, 60 s at 72°C), and a final elongation step (10 min at 72°C). The multiplex PCR products were prepared as described above. The specificity of the multiplex PCR was confirmed by comparing the amplicons obtained by multiplex PCR with those obtained by means of single PCRs whose products were secondarily pooled for capillary electrophoresis. The sizes of the amplicons were determined by using an internal size standard (GeneScan-500 ROX; Applied Biosystems), as described above.

RESULTS

Short microsatellite distribution in the complete sequence of HSV-1 genome.

The in silico analysis by use of the MsatFinder algorithm revealed that the HSV-1 genome (strain 17, GenBank accession number X14112) contained at least 79 targeted short repeat sites (arbitrarily numbered HSV1-M1 to HSV1-M79), including 35 mononucleotide repeat units, 16 dinucleotide repeats, and 28 trinucleotide repeats (Fig. 1). As expected, they were mainly localized in noncoding sequences, including terminal repeat regions (TR_L, TR_S) and inverted internal repeat regions (IR_L, IR_S) as well as intergenic regions, while 34 were in open reading frames (Fig. 1). The maximum number of repeat units of dinucleotide or trinucleotide motif was 6 [M33-(GC)6 and M41-(CGT)6], while the longest mononucleotide tract was 15 bases [M70-(C)15] (Table 2). Among these microsatellites, 29 were selected and 25 primer pairs were designed in order to synthesize amplicons containing these target repeats (Table 1).

FIG. 1.

Linear map of genomic distribution of HSV-1 microsatellites. The linear double-stranded DNA is represented with the nucleotide scale at the bottom according to the sequence of strain 17 (GenBank accession number X14112). The unique U_L and U_S sequences are shown as heavy solid lines, and the terminal and internal inverted repeats (TR_L, IR_L, IR_S, and TR_S) are shown as hatched boxes. The open reading frames on both the forward and the reverse strands are represented by open arrows. Each vertical line represents one microsatellite locus; the dotted lines and the dashed lines represent monomorphic and polymorphic microsatellites in the samples tested, respectively. This map is restricted to sites with at least four trinucleotide, five dinucleotide, or nine mononucleotide repeats units.

TABLE 2.

Determination of microsatellite haplotypes for strain KOS and the 21 clinical isolates studied

Straina	Site	Date (mo/day/yr)	Length (bp) of the following allele:
			M30	M10	M14	M69	M22	M41	M26	M23	M31/32	M5
KOS			162	170	225	245	296	366	371	400	456	471
Clinical isolates
R	Face	11/20/2006	137	174	228	245	297	366	371	403	447	469
K	Throat	09/12/2005	138	172	229	242	294	366	374	400	459	471
B	Throat	10/10/2005	138	174	225	242	303	363	370	408	450	469
G	Throat	05/12/2005	138	174	226	245	294	366	372	400	459	470
L1	Lip	10/17/2005	138	175	226	245	297	366	371	401	459	469
L2	Cornea	12/14/2005	138	175	226	245	297	366	371	401	459	469
F	Throat	04/15/2005	139	171	224	245	296	366	371	400	460	470
I	Throat	04/13/2005	139	172	223	242	297	363	370	406	452	474
N1	Mouth	10/03/2007	140	170	224	245	295	366	374	406	450	473
N2	Mouth	11/10/2007	140	170	224	245	295	366	374	406	450	473
O1	Tongue	04/23/2007	140	171	227	245	294	363	369	407	448	474
O2	Lip	05/18/2007	140	171	227	245	294	363	369	407	448	474
H	Lip	01/14/2005	140	172	225	245	297	363	372	400	458	473
D	Lip	04/20/2005	140	173	224	242	296	366	371	410	448	469
A	Throat	01/10/2005	140	173	225	242	303	363	370	411	452	470
J	Throat	11/12/2005	141	166	223	242	293	363	374	405	449	476
C	Throat	12/22/2004	147	174	224	242	295	363	369	403	456	472
E	Throat	04/04/2005	149	173	226	242	297	366	370	411	457	471
P1	Nose	03/07/2007	157	172	229	245	296	366	371	409	451	475
P2	Lip	03/19/2007	157	172	229	245	296	366	371	409	451	475
Q	Genitals	11/20/2006	158	170	224	245	295	366	374	406	449	472

^aStrain pairs L1-L2, N1-N2, O1-O2, and P1-P2 were obtained from the same patient at two different times.

Characterization of HSV-1 microsatellites.

We first surveyed the 29 selected microsatellites in strain KOS and 11 HSV-1 clinical isolates (isolates A to K) using the 25 different PCR primer pairs separately (Table 1). The study of the corresponding amplified fragments by means of capillary electrophoresis showed that 19 of these loci were polymorphic, with microsatellites M24, M28, M33, M36, M39, and M42 being monomorphic among the panel of HSV-1 isolates tested. The polymorphic alleles are presented in Table S1 in the supplemental material. At the same time, the variability of each microsatellite was confirmed by sequencing methods. As expected, mononucleotide repeats were more variable, with 5 to 7 variants detected among the 12 strains tested. M21, M39, and M51 were considered impure microsatellites since they presented polymorphisms in both their lengths and sequences. Polymorphic trinucleotide microsatellites were found within two protein-coding genes: M41(CGT)6, which encodes a stretch of aspartic acid residues within the UL46 gene, which encodes the tegument protein, and M69(GGA)4, which codes for a polyglutamyl stretch within US4, which encodes virion gG (Table 1). At these two sites, two variant sequences were found among all strains tested in this study (Fig. 2). From this preliminary analysis, it was concluded that short microsatellites corresponding to mono-, di-, and trinucleotide repeats were highly polymorphic in the HSV-1 genome and can be considered potential molecular markers for the precise differentiation of HSV-1 strains.

FIG. 2.

Sequences of two polymorphic trinucleotide microsatellites present in HSV-1 genes. (A) Microsatellite analysis of M41(CGT)6, which codes for a stretch of aspartic acid residues (residues 582 to 587) within the UL46 gene, showed two variants, M41(CGT)6 and M41(CGT)5, that led to alleles 366 and 363, respectively. The two variants represented 70% and 30% of the HSV-1 strains tested in this study, respectively. (B) Microsatellite analysis of M69(GGA)4, which codes for a polyglutamyl stretch (residues 79 to 82) within the US4 gene showed two variants, M69(GGA)4 and M69(GGA)5, that led to alleles 242 and 245, respectively. The two variants represented 42% and 58% of the HSV-1 strains tested in this study, respectively.

Microsatellite haplotypes obtained by multiplex assay for different HSV-1 strains.

Ten primer pairs amplifying polymorphic sites that could be multiplexed in the same PCR run were chosen in order to facilitate the rapid differentiation of HSV-1 strains by their microsatellite haplotypes (Table 1). In the present study, the haplotype was defined as the combination of the lengths of the 10 PCR products obtained from the multiplex PCR assay and analyzed by polyacrylamide capillary electrophoresis (Fig. 3). The optimization of the multiplex reaction conditions led to the final definition of the technique described in Materials and Methods. The interassay reproducibility of the multiplex PCR was assessed by amplifying sample A in 50 independent reactions, and no variation in the haplotype was observed. The microsatellite haplotypes of the 21 clinical isolates and strain KOS are presented in Table 2. As expected, the haplotypes corresponding to distinct HSV-1 isolates obtained from different individuals were all different. Considering the number of variants described for each of these 10 microsatellites, it was computed that this strategy would be able, in theory, to identify more than 10⁷ different haplotypes. Therefore, the multiplex assay was found to be simple, reproducible, and highly discriminative.

FIG. 3.

Electropherograms of multiplex PCR products illustrating the haplotype characteristics used to differentiate the HSV-1 strains. The haplotype was defined as the combination of the lengths of 10 amplicons labeled at the 5′ end with 6-carboxyfluorescein and containing HSV-1 microsatellite sequences obtained from the multiplex assay and analyzed by polyacrylamide capillary electrophoresis. The numbers 100 to 500 indicate the sizes (in base pairs) of the PCR amplicons. The haplotypes of strain B (top panel) and strain F (bottom panel) represented here display the length polymorphisms for the 10 loci.

The stability of the microsatellites was investigated by prolonged serial propagation of HSV-1 in cell cultures. Strain KOS and four clinical isolates (isolates A, B, C, and F) were plaque purified three times and were then serially propagated for 15 passages on Vero cells. In order to take the impact of the cellular environment into account, we also used different cellular systems. For that purpose, two strains (strains KOS and A) were propagated in separate trials: on Vero cells alone, on human MRC5 fibroblasts alone, and under a third condition that used the repetitive rotation of Vero and MRC5 cells. At each of the 15 passages and in the contexts of the different host cells, the microsatellite haplotypes of the strains were analyzed. No variation was observed, with each of the five strains keeping its own haplotype, depicted in Table 2, at each passage (data not shown). These data suggest that the HSV-1 microsatellites remained stable during viral replication. This led us to conclude that the multiplex PCR and length polymorphism analysis could be used as a rapid and precise tool for the identification of HSV-1 strains.

Differentiation of clinical HSV-1 isolates.

The microsatellite haplotypes of two serial isolates from the same individual were compared (Table 2). Isolate L1 was isolated from a patient's lip vesicles; isolate L2 was isolated 59 days later from the same patient's herpetic keratitis lesions. Isolates N1 and N2 were isolated 39 days apart from a patient with a stem cell allograft who presented with chronic HSV-1 infection of the oral mucosa. Isolates O1 and O2 were isolated from the same individual from two different herpetic recurrence lesions on the tongue and on the inferior lip, respectively. Isolates P1 and P2 came from a patient who developed herpetic vesicles 2 weeks apart, first on the skin around the nose and then on the lip. The microsatellite haplotype was identical in each case, even though the locations of the herpes lesions slightly differed over time. These results indicate that in these cases the strain haplotype remained stable in vivo through distinct or chronic herpetic episodes in the same individual.

Finally, strains Q and R were from two different patients attending the same care unit. The first patient, an immunosuppressed 80-year-old woman, was hospitalized with a case of severe genital herpes due to HSV-1 that required intravenous acyclovir treatment. The second patient was hospitalized for acute eczematous lesions of facial atopic dermatitis and who developed a HSV-1 superinfection at exactly the same time. While this might have suggested the transmission of HSV-1 by hand carriage from the first patient to the second, the haplotypes of the two isolates were very different, thus indicating, on the contrary, that the infections originated from different viral strains (Table 2).

DISCUSSION

This study reports for the first time the presence of numerous microsatellite loci across the HSV-1 genome, found mostly, as expected, in noncoding regions and characterized by a high level of length polymorphism. Short microsatellite polymorphisms have already been used for the precise characterization of HCMV strains (9, 29, 51), and a similar proposal has been made for human herpesvirus 6 strains regarding the specific copy number of telomeric repeat sequences located at the ends of the genome (25). In comparison with reiterated regions in the U_S component still currently used for HSV-1 strain differentiation (33), short microsatellites have the advantage of being distributed across all parts of the HSV-1 genome. The HSV-1 microsatellites remain stable for any given strain after serial propagation in cell culture in vitro and in vivo. The results of this preliminary longitudinal study indicate that sequential HSV-1 isolates obtained from one individual present the same microsatellite profiles. Moreover, the multiplex PCR-based analysis of microsatellites presented here is a simple, specific, precise, and reproducible method that rapidly provides results that can be used to compare HSV-1 strains. Altogether, these results establish the potential use of microsatellite polymorphism analysis in epidemiological studies of HSV-1 infections.

As far as the mechanism of microsatellite variability is concerned, the changes in repeat numbers at microsatellite loci are generally caused by slip-strand mispairing errors during DNA replication (42). It has been recognized that in prokaryotic cells, as well as in eukaryotic cells, the integrity of the genome is under the control of complex DNA repair systems, including the mismatch repair system. Their functions are used to recognize and repair erroneous insertions, deletions, and the misincorporation of bases that can arise during DNA replication and recombination, as well as to repair some forms of DNA damage (2, 15, 50). It is known that some mutational events are normally proofread during replication by the HSV-1 DNA polymerase, which possesses an intrinsic 3′ to 5′ exonuclease activity that attends to the removal of incorrectly incorporated deoxyribonucleoside triphosphates (41). Furthermore, during HSV-1 infection, some host cell DNA repair proteins are recruited to nuclear replication compartments (20, 52). The recruitment of host DNA repair proteins seems to benefit viral replication, but their actual roles need to be confirmed. The influence of the recruitment of the mismatch repair system by HSV-1 on the stability or instability of its microsatellite sequences has not yet been studied. Additionally, some authors have demonstrated that the reduced DNA binding of the UL42 processivity subunit of the HSV-1 DNA polymerase is associated with a significant decrease in replication fidelity (18). All these elements could underlie the diversity of HSV-1 microsatellites observed in our study.

Regarding their impact on the physiology of HSV-1, microsatellites might exert subtle influences on the regulation of DNA metabolic processes and viral gene expression as well as chromatin organization, as shown in eukaryote and prokaryote genomes. There is abundant evidence indicating that changes in gene sequence length could inactivate, affect, or change protein function, as shown in human cancer and genetic disorders and by the modification of pathogenesis in bacteria (19). Microsatellites may provide a molecular basis for fast adaptation to environmental changes or selective pressure. For instance, many of the mutations associated with the acyclovir-resistant HSV-1 isolates contain insertions or deletions in short homopolymeric G or C sequences in the thymidine kinase (TK) gene (40). Some of these regions are remarkably instable and likewise give to the TK-negative mutants the possibility to revert to a TK-positive phenotype, while the phenotype gives the virus the advantage of being able to reactivate and grow in ganglia (13). These variations in the length of the sequence of the TK gene of HSV-1 illustrate the adaptive mechanisms that permit the virus to evade acyclovir therapy yet retain its pathogenicity. In our study, the screen that identified 79 microsatellites does not yet constitute an exhaustive survey of the HSV-1 genome; we looked only for total repeats longer than the chosen cutoff (microsatellites of four trinucleotides, five dinucleotides, or nine mononucleotides). This cutoff explains why homopolymeric G or C sequences in the TK gene were not included. In our study, two polymorphic trinucleotide microsatellites were found within two protein sequences. This may result in changes in the UL46 tegument protein, which modulates the transcriptional induction of alpha genes (3, 54), and US4, which encodes gG, which is a viral envelope glycoprotein that is assumed to contribute to viral entry through the apical surfaces of polarized cells (46). In the second case, our results confirmed the sequence variations of the GAG tandem repeat flanking GAA in gG that were already reported by others (27, 30). To date, these amino acid sequence modifications and their biological significance have not yet been reported.

Otherwise, although the effects of polymorphisms in microsatellites on the physiology of HSV-1 are largely unknown, the fact that microsatellites were more often found in noncoding regions than in open reading frames suggests that they have no influence on either the virulence of the strain or the evolution of the virus. Since these variations must likely occur at random without offering any replicative or adaptive advantage, these microsatellites may thereby be considered neutral molecular markers. In every instance, they might be used in studies in which HSV-1 strain tracing is necessary: in physiological and epidemiological studies (34) or when nosocomial HSV-1 infections (21, 38) or reinfection with a novel strain (31) are suspected. Our study has indentified 23 polymorphic microsatellite loci that may be used as molecular markers for epidemiological analysis of HSV-1 strains and HSV-1 strain differentiation. Furthermore, microsatellite-based phylogeny and principal-component analyses, which are the methods usually described in the literature, might be applied to show the clustering of HSV-1 strains. Microsatellites accurately characterized the HSV-1 strains from patients evaluated in the present study and were used to ascertain the identity between strains isolated sequentially from the same patient, as well as to differentiate strains in the case of the hypothetical nosocomial transmission of HSV-1 strains. Our preliminary results are promising and should have implications for further epidemiological studies of the transmission and the evolution of HSV-1 strains.