- Email: [email protected]
Development of a point mutation assay for the detection of human cytomegalovirus UL97 mutations associated with ganciclovir resistance
Journal of Virological Methods
ELSEVIER
Journal of Virological Methods 68 (1997) 225 234
Development of a point mutation assay for the detection of human cytomegalovirus UL97 mutations associated with ganciclovir resistance E.
Frances Bowen a, Margaret A. Johnson b, Paul D. Griffiths a, Vincent C. Emery
a,,
a Department of Virology Royal Free Hospital School of Medieine, Rowland Hill Street, London N W 3 2QG, UK b AIDS Unit, Royal Free Hospital School of Medicine, Rowland Hill Street, London N W 3 2QG, UK
Received 26 March 1997: received in revised form 29 July 1997; accepted 29 July 1997
Abstract A point mutation assay was developed to detect the quantitative prevalence of mutations at codons 460 (M to I; M to V), 520 (H to Q), 594 (A to V) and 595 (L to F: L to S) within the UL97 gene of human cytomegalovirus which segregate with ganciclovir resistance. Synthetic mixtures of wild-type and mutant plasmids containing the UL97 gene were amplified by nested polymerase chain reaction and the 700 base pair amplicon subsequently subjected to the point mutation assay. In plasmid reconstruction experiments, there was a high correlation between experimentally derived percentage mutant with the theoretical values. The assay was then used to assess the changes in the genetic composition of the UL97 gene in three patients on prolonged ganciclovir therapy. All three patients developed genotypic resistance against ganciclovir involving mutation at codon L595S, L595F and double mutation at codons L595F and M460I. In one patient, alteration of therapy to foscarnet did not affect the composition of UL97 and virus remained genotypically resistant to ganciclovir. In contrast, in two patients whose therapy was altered to cidofovir (HPMPC), repopulation with cytomegalovirus strains carrying the wild-type (ganciclovir-sensitive) codon at positions 595 and 460 occurred. The potential use of this assay for the rapid detection of cytomegalovirus resistance in patients on long-term ganciclovir therapy is discussed. © 1997 Elsevier Science B.V. Keywords: Cidofovir; Resistance; Therapy; Dynamics
1. Introduction
* Corresponding author.
Since 1989, ganciclovir (GCV) has been the first line treatment for h u m a n cytomegalovirus ( H C M V ) disease in patients with Acquired Ira-
0166-0934/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PH S0166-0934(97)00131-6
226
E.F. Bowen et al. /Journal of Virological Methods' 68 (1997) 225 234
munodeficiency Syndrome (AIDS) (reviewed by Crumpacker, 1996). This widespread usage has led to the emergence of GCV resistant strains of HCMV that are becoming more problematic as patients with HCMV retinitis are living longer and hence receiving longer courses of ganciclovir. In addition, with the results of the oral ganciclovir study 1654 there is likely to be increasing use of prophylactic GCV (Spector et al., 1996). The frequency of resistant strains has been estimated at 7.6% of patients receiving GCV for more than 3 months (Drew et al., 1991). However, it should be noted that these resistant strains were identified in urine samples after passage in vitro and may not truly represent resistance patterns in vivo. In 1992, two elegant molecular studies demonstrated that the HCMV UL97 open reading frame was responsible for the monophosphorylation of ganciclovir (Littler et al., 1992; Sullivan et al., 1992). This discovery has allowed the development of molecular techniques such as PCR to amplify UL97 both from cell culture propagated strains of HCMV derived from patients or directly from clinical samples. The UL97 DNA sequence in clinical samples (predominately plasma) has been determined by direct sequencing methods using 33p end-labelled primers, the incorporation of 35Slabelled dATP (Chou et al., 1995a; Hanson et al., 1995; Wolf et al., 1995) or silver staining of the sequencing products (Alain et al., 1995). From these studies, the most frequently identified mutations in clinical samples obtained from patients on prolonged GCV therapy are at amino acids M460V, M460I, H520Q, A594V, L595S and L595F (Lurain et al., 1994; Alain et al., 1995; Chou et al., 1995b; Hanson et al., 1995; Wolf et al., 1995; Baldanti et al., 1996; Boivin et al., 1996). All of these mutations have been shown by marker transfer studies to confer GCV resistance to Ad169. Five of these mutations (V460, Q520, V594, $595 and F595) can be detected by a combination of PCR followed by restriction endonuclease digestion (Chou et al., 1995b; Hanson et al., 1995). Although these methods potentially allow the rapid identification of resistant virus in clinical samples, they do not provide quantitative information of the prevalence of these mutations within the viral population. Such data will be
important in monitoring the evolution of GCV resistance of HCMV in vivo and to allow modifications to therapy at appropriate times. In order to gain insight into the quantitative prevalence of GCV resistance genotypes in patients on prolonged GCV therapy, a rapid point mutation assay (PMA) was developed to screen for the mutations M460V, M460I, H520Q, A594V, L595S and L595F and used to determine the quantitative prevalence of UL97 mutations in multiple samples from AIDS retinitis patients receiving prolonged exposure to anti-HCMV therapy.
2. Materials and methods 2.1. Patients
A subset of HIV-seropositive patients who attended the AIDS clinic at the Royal Free Hospital from 1 September 1993 with a diagnosis of cytomegalovirus retinitis were used to assess the performance of the point mutation assay (Darlington et al., 1991; Bowen et al., 1996). Samples of blood (4 ml in citrate filled tubes) and urine were collected before treatment (on the day of diagnosis of retinitis), at the end of the induction period, and then monthly thereafter. Extra samples of blood and urine were collected at times of retinitis progression. All patients received ganciclovir for induction therapy and as primary maintenance therapy either intravenously or orally. Intolerance to ganciclovir or recurrent progression of HCMV retinitis was treated with either foscarnet or cidofovir. Three patients in whom substantial numbers of blood samples ( > 10) were available were selected for detailed analysis with the point mutation assay described below. 2.2. Polymerase chain reaction
Initially all samples were subjected to a qualitative PCR for the HCMV gene gB (Darlington et al., 1991; Bowen et al., 1996). All HCMV positive samples were then quantitated using a quantitative-competitive PCR previously described (Fox et
E.F. Bowen et al./Journal of Virological Methods 68 (7997) 225 234 al., 1992, 1995; Bowen et al., 1996) and analysed for UL97. A nested PCR for the detection of UL97 was developed using D N A extracted from whole blood using a commercial ion-exchange column (Qiagen, UK). The outer primers were as follows: 1. UL97 A 5' A G A C G G T G C T A C G G T C T G G A T G T 3' 2. UL97 B 5' G T T T G T A C C T T C T C T G T T G C C T T T 3' Each PCR reaction contained 150 ng primer, 3 m M MgC12, 200 # M of each dNTP and 2 units of Taq polymerase in 10 x PCR buffer containing 250 m M Tris-HC1 pH 8.4, 170 m M ammonium sulphate, 100 m M fl-mercaptoethanol, 0.02% (w/ v) gelatin. Following the addition of 5 /~1 target D N A the reaction was made up to a final volume of 50 ill with SDW. Cycling conditions were as follows; 95°C for 6 min followed by 25 cycles of 94°C 1 min, 45°C 2 min, 72°C 3 min; 2 tL1 of the outer PCR product was then amplified in a nested PCR. In each nested PCR reaction one of the primers was 5' biotinylated to enable adherence of one D N A strand to the wells of a streptavidincoated microtitre plate. The nested primer sequences were as follows: 1. UL97 C 5' C A A C G T C A C G G T A C A T C G A C GTTT 2. UL97 D 5' G C C A T G C T C G C C C A G G A G A CAGG In each PCR, 50 ng of biotinylated primer (either primer C or D depending upon the specific P M A probe used; see later) was used in conjunction with 100 ng non-biotinylated primer. The inner round PCR conditions were the same as described for the outer-round PCR except for the use of 1.5 m M MgC12 in a total reaction volume of 100 itl. Cycling conditions were 35 cycles of 94°C 1 rain, 58°C 2 min, 72°C 3 rain. The presence of the 700 bp amplicons were detected by electrophoresing 10 /~1 PCR product on a 1% agarose gel in the presence of a 2 HindIII/EcoR1 D N A marker. 2.3. Design of oligonucleotide probes The oligonucleotide probes were designed to bind to the biotinylated strand of the PCR
227
product and terminated one base prior to the point mutation to be detected. The sequences for the probes were taken from the published sequence of either the sense or anti-sense strand of Ad169 (Chee et al., 1990). In areas of the UL97 gene where common silent polymorphisms were known to occur in clinical samples a mixed base was included in the probe sequence. All oligonucleotide probes were synthesised and H P L C purified by Cruachem, UK. Probes were made to detect the six most common mutations; M460V (ACT-to-GTG), M460I (ATG-to-ATT), H520Q (CAC-to-CAG), A594V (GCG-to-GTG), L595S (TTG-to-TCG) and L595F (TTG-to-TTT). Probes I460, V594, $595 and F595 were complementary to the biotinylated sense strand of the PCR product whilst probes V460 and Q520 bound to the biotinylated anti-sense strand. The probe sequences were as follows: 1. Probe 1: L595F; 5' A G T G C G T G A G C T T A / G C C G T T C T C antisense ~ 5'B UL97C 2. Probe 2: L595S; 5' G T G C G T G A G C T T A / G C C G T T C T C C antisense --* 5'B UL97C 3. Probe 3: A594V; 5 ' G T G A G C T T A / G C C G T T CTCCAAC antisense --, 5'B UL97C 4. Probe 4: H520Q; 5' C C G T C T G C G C G A A T G T T A C C A sense -~ 5'B UL97D 5. Probe 5: M460I; 5' G T T C A C G T C G A T G A G C A C G T T antisense --, 5'BUL97C 6. Probe 6: M460V; 5' T G C C A C T T T G A C / T A T T A C A C C C sense ~ 5'B UL97D where 5'B indicates the presence of biotin on either primer C or D. 2.4. Microtitre point mutation assay The microtitre point mutation assay (PMA) to detect H C M V UL97 mutations was based on the approach originally described by Kaye et al. (1992) for the detection of zidovudine specific mutations in the R T gene of HIV-1. Minor modifications to optimise the performance of this assay for UL97 were effected. These included the use of 96-well streptavidin-coated plates produced by Advanced Biotechnologies Ltd., optimisation of the binding of the biotinylated PCR product to the plate and the use of a lower annealing temperature during the hybridisation of the probe to the biotinylated strand.
228
E.F. Bowen et al./Journal of Virological Methods 68 (J997) 225 234
Following the nested PCR, 10 /~1 of biotinylated PCR product were diluted to 25/~1 with 1 x T T A buffer (10 x TTA; 100 m M Tris-HC1 pH 7.6, 0.5% Tween 20, 1.0% sodium azide stored at room temperature) and added to each of four microtitre wells of a 96-well streptavidin-coated plate (400 ng affinity purified streptavidin/300 Itl well; StrepMax, Advanced Biotechnologies Ltd.) and incubated at 37°C for 60 rain. The wells were washed three times with TTA buffer and then 40 ¢tl 0.15 M N a O H were added to each well and left to incubate for 5 rain at room temperature to denature the captured PCR product. The wells were washed four times with T T A to remove the released second strand and 25/~1 of annealing mix made up of the appropriate oligonucleotide probe in 2.5 ml P M A diluent (PMA diluent; 40 mM Tris HC1 pH 7.6, 20 mM MgC12, 50 m M NaC1 stored at 4°C) was added to each well. To optimise the PMA for UL97, oligonucleotide probe concentrations from 3.3 ng to 26 ng were used to maximise probe binding to the biotinylated product whilst keeping non-specific binding to a minimum. The plate was covered and placed in a 55°C water bath for 3 rain and then allowed to cool slowly to room temperature over at least 30 min; 24/~1 of each of the four 35S-labelled dNTPs (1200 Ci/mmol (NEN DuPont) stored at - 2 0 ° C as a 1 in 10 dilution in PMA diluent) was added to a 168-/~1 aliquot of 0.1 M DTT. To this mixture was then added Klenow polymerase (10 units, Boehringer Mannheim), diluted in 200 /~1 P M A diluent. A 9.8-~1 aliquot of the resultant labelling mix was added to each well with a different dNTP being added to each of the four wells, and incubated at room temperature for 3 min. The wells were washed four times with TTA to remove unincorporated label from the target strand and 40 i~1 of 0.15 M N a O H was then added to denature the probe and any incorporated labelled nucleotide. The N a O H solution was removed and mixed with 5 ml scintillation cocktail (Optiphase Hisafe-3) in 10-ml scintillation vials (Camlab, Cambridge, UK). The radioactivity in the samples was determined in a LKB Rackbeta scintillation counter.
2.5. Generation o f the standard curve f o r mixtures o f wild-type and mutant codons
In order to establish a standard curve for the differentiation of mutant and wild-type sequences by the PMA, standard amounts of UL97 D N A of a known sequence were amplified by PCR and subjected to the PMA. For these experiments three separate UL97 clones derived by PCR amplification of blood from patients on long-term GCV therapy (which had been previously analysed by conventional D N A sequencing methods of cloned UL97 amplicons) were used. Each clone possessed a single mutation at codons 594 (A to V), 595 (L to S) and 595 (L to F). In each PCR the ratio of the two genotypes was mixed as follows; 100% WT, 75% WT: 25% MUT, 50% WT: 50% MUT, 25% WT: 75% MUT, 100% MUT. Each clone was used as either WT or M U T depending on the probe used. For example, when mixing A594V and L595S, probe 3 detects A595V as M U T and L595S as WT, whilst probe 2 detects L595S as M U T and A595V as WT. Each reaction was performed in triplicate. Probes 4 - 6 were standardised on clones that were known to contain wild-type sequences at that particular mutation. Linear regression fitting was used to determine the adherence of the data points to a straight line.
3. Results 3. I. Generation o f the standard curves
The ability of the optimised point mutation assay described in the methods section to detect different ratios of wild-type or mutant nucleotides at specific codons was assessed by analysis of pre-determined mixtures obtained from previously cloned and sequenced UL97 products derived from clinical samples. The data from three independent experiments were used to construct standard curves for L595F, L595S and A595V relating predicted ratios to wild-type: mutant to the experimentally determined ratios. The results are shown in Fig. 1 together with the linear regression line-fit. In each PMA the experimentally deter-
E.F. Bowen et al./Journal of Virological Methods 68 (1997) 225 234 100 -
100 -
80
80
so
_~ 6O ._o
~ 40
u_ 40 U3
229
03
20 P = O" 0002
0 2
0
'
I
20
'
I
'
40
I
'
I
60
'
20
--1-
0 :I 0
I
80
-
100
'
I
P
'
20
I
40
% 595 S (theoretical)
'
I
'
60
I
80
=0"0008
'
I
100
% 595 F (theoret!cal)
100 -
80
_~ 60 o > 40 O'1 Lf)
~0.~/
P = 0.0001
z/ '
0
I
20
'
l
40
'
I
J
60
l
80
'
I
100
% 594 V (theoretical)
Fig. 1. Calibration curves relating the experimentally determined ratio of wild-type and mutant genotypes at codons L595F, L595S and M460V to the predicted ratio. Each data point was generated from three separate PMAs and the mean + 1 standard deviation are quoted on the graph. The line of best fit was computed using the line fitting program available with Microcal Origin. The adherence of line to the data is given by the R-value and its associated significance indicated. m i n e d r a t i o o f w i l d - t y p e to m u t a n t was in close a g r e e m e n t to the p r e d i c t e d ratios.
3.2. Analysis o f clinical samples Jor the presence o f UL97 mutations I n o r d e r to ascertain w h e t h e r the P M A c o u l d be used successfully on clinical samples, m u l t i p l e sam-
ples were a n a l y s e d f r o m three p a t i e n t s w h o develop e d H C M V retinitis a n d were initially t r e a t e d with ganciclovir t h e r a p y , then with either f o s c a r n e t o r c i d o f o v i r due the a p p a r e n t clinical failure o f ganciclovir. It was r e a s o n e d t h a t these p a t i e n t s were likely to have ganciclovir resistant strains o f H C M V on the basis o f their clinical h i s t o r y a n d so w o u l d p r o vide suitable samples on which to assess the assay.
230
E.F. Bowen et al./Journal of Virological Methods 68 (1997) 225-234
Each blood DNA extract was subjected to the PMA for the five most common coding mutations that have been shown to endow ganciclovir resistance to Ad169 (see above). In all cases, pre-treatment samples were also analysed by the PMA. The results are shown in Table 1 and indicate the percentage of mutant virus present at codons 595, 594, 520 and 460 for each patient at multiple time-points. The anti-HCMV therapy in use at the time the sample was obtained is also included. In patients 1 and 2, the PMA identified the presence of a low but reproducible amount of the L595S and L595F mutations respectively at the diagnosis of HCMV retinitis, i.e. before either patient had been exposed to ganciclovir. In contrast the PMA results for patient 3 indicated only wild-type sequences at all codons. During ganciclovir therapy, the proportion of L595F and L595S genotypes in patients 1 and 2 increased rapidly to reach a level greater than 90% by days 179 and 111 post-diagnosis of retinitis respectively. The substitution of cidofovir for ganciclovir in patient 1 at day 279 resulted in a gradual repopulation with HCMV strains that were wildtype at codon 595. In patient 2, whose therapy was altered to foscarnet at day 272, no substantial alterations in the proportion of L595F genotype were observed. In this patient, HCMV was detected in the urine at day 308 and was predominately wild-type at codon 595 (95%) despite the presence of predominately mutant virus in the blood. However, mutant virus at this codon rapidly repopulated the virus strains and within 3 weeks of GCV therapy for gastrointestinal disease was approximately 90% mutant at codon 595. The genotypic resistance profile observed in patient 3 is more complex. Both first and second clinical progressions of retinitis were not accompanied by detection of HCMV in the blood by PCR. However, at day 198 virus was detected which was 78% mutant at codon 595 (L to F) and 32% mutant at codon 460 (M to I). Following a short period of foscarnet the patient was initiated on cidofovir therapy. Similar to the results obtained for patient 1, the repopulation of wild-type UL97 genotypes occurred at both the L595F and M460I codons. Viral load in all three patients remained elevated following the development of genotypic resistance to ganciclovir in UL97 despite the intervention with
drugs that do not require UL97 for activation. A ty-pical graph of the viral load profile and alterations in the genotypic composition of UL97 is shown in Fig. 2.
4. Discussion
The development of a point mutation assay is described which provides a rapid and reproducible means of analysing genetic changes in UL97 amplified directly from clinical samples. The method allows the prospective study of changes in UL97 under the influence of different anti-HCMV drugs without the need for in vitro passage. To our knowledge this is the first time that quantitative prevalence of mutations in patients undergoing antiHCMV therapy has been detailed. The assay was similar with respect to reproducibility and ease of use to that previously described for HIV RT mutations segregating with AZT resistance (Kaye et al., 1992). Prospective analysis of multiple samples derived from three patients treated with GCV for HCMV retinitis using the PMA revealed the presence of one or more mutations in codons 595 and 460 of the UL97 gene which are known from other studies to confer reduced susceptibility to GCV. In these patients the availability of multiple PCR positive follow-up samples enabled the evolution of GCV resistance at the UL97 locus to be examined and the influence of alternations in therapy on the re-emergence of UL97 wild-type strains investigated. In patient 1, the virus was > 90% mutant at codon 595 at the time of resurgence of PCR viraemia (day 179) and the first clinical progression of the retinitis was observed at day 241. Due to the failure of this individual to respond to GCV re-induction therapy and evidence of further clinical progression the patient was switched to cidofovir therapy at day 279. Over the ensuing 151 days outgrowth of the wild-type UL97 (L595) virus occurred. This observation is consistent with the fact that cidofovir inhibits CMV replication through the DNA polymerase and hence UL97 mutants have no selective growth advantage under the new circumstances. Despite the antiviral effects of cidofovir, the viral load in this patient remained
E.F. Bowen et al. /Journal of Virological Methods 68 (1997) 225-234
231
Table 1 Distribution of genotypes at codons L595F, L595S, A594V, H520Q, M460I and M460V in the UL97 gene of H C M V in multiple blood samples from three patients by the point mutation assay Pt
HCMV Diagnosis
Days
Drug
1
Retinitis
0 179 241 248 274 279 301 329 361 375 403 430
-GCV GCV GCV GCV CDV CDV CDV CDV CDV CDV CDV
wt wt wt wt wt wt wt wt wt wt wt wt
0 93 111 127 134 180 204 225 272 308 322 397 438 468
GCV GCV GCV GCV GCV GCV GCV PFA PFA PFA PFA PFA PFA
9% 79% 95% 99% 99.1% 97% 99.6% 97% 98% 97% 97% 97% 95% 88%
0 77 133 198 230 247 280 296 301 320 330
GCV GCV GCV GCV PFA CDV CDV CDV CDV CDV
wt -ve - ve 78% 79% 77% 81% 78% 66% 24% 2%
1st R P 2nd R P
2
Retinitis 1st R P
2nd R P 3rd R P
3
Retinitis 1st R P 2nd R P
3rd R P
L595F (%F)
L595S (%S)
8%
4% - ve - ve - ve -re - ve -ve - ve - ve 5% N/A 87% 90%, 92%
A594V (%V)
H520Q (%Q)
M460I (%I)
M460V (%V)
94% 97% 96% 95% 90.5% 70% 61% 19%, 3% 7% 8%
wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt -ve - ve wt wt wt wt wt wt wt wt
wt -ve - ve wt wt wt wt wt wt wt wt
wt -ve - ve wt wt wt wt wt wt wt wt
wt -ve - ve 32% 84% 93% 98% 98.5% 98% 22°/,, 4%
wt -ve - ve wt wt wt wt wt wt wt wt
In patient 2, the % m u t a n t at codon L595F is quoted for blood (left-hand data set) and urine (right-hand data set). R P = retinitis progression; G C V = ganciclovir, P F A = foscarnet; C D V = cidofovir, wt = wild-type at codon of interest. high
implying
that
resistant to GCV polymerase
CMV
strains
may
be dually
and cidofovir at the UL54 (DNA
l o c u s ) w h e r e p r e s e n t . T h i s p o s s i b i l i t y is
currently
being investigated.
observed
in patient
evolution of HCMV
A similar profile was
3, e x c e p t
that
in this
case
strains carrying two mutations
in UL97 were observed
(codons
595 and 460), both
of which
reverted
to wild-type
of cidofovir. Evolution was
not
between
parallel
following
exposure
of resistance at each codon
implying
differences
viral strains carrying
in
fitness
single mutations
codons 595 and 460 or dual mutations. of cidofovir to inhibit systemic HCMV
at
The inability replication
in spite of its effects at stabilising HCMV
retinitis
E.F. Bowen et al. /Journal of Virological Methods 68 (1997) 225-234
232 HCMV
RETINITIS
RP
RP
RP
--
100
80
60 % 595 S or 4601
HCMV log10 genomes/ml blood
40
20
0 ,
0
100
200
300
I
400
500
DAYS
GCV Induction
RP
RETINITIS PROGRESSION
GCV Maintenance
[]
HCMV load
Intravitreal FOSCARNET
•
% 595 S
CIDOFOVIR
O
% 460 I
Fig. 2. Longitudinal viral load profiles and alterations of UL97 genotypic resistance in patient 3 of Table 1. Periods and type of therapy are indicated together with clinical progression of HCMV retinitis.
has been recently reported (Lalezari et al., 1997; SOCA, 1997). In patient 2, evolution of high level GCV resistance at UL97 was observed over a period of 111 days which did not alter following exposure of the patient to foscarnet. Since foscarnet also targets the UL54 gene product of HCMV directly, it is intriguing why the viral species was not repopulated by strains carrying the wild-type codon at position 595. Since viral load remained elevated in this patient during foscarnet therapy it is plausible that this individual harboured DNA polymerase mutants resistant to GCV and foscarnet. Again this
possibility is currently under investigation. The temporal dislocation of virus detection in different body sites is aptly demonstrated in this patient. Despite the presence of GCV resistance virus in blood at high viral loads for over 200 days, virus was detected in urine relatively late in the clinical course. Initially, this virus was wild-type at codon 595 but the population was rapidly re-populated by a strain carrying the same GCV resistant mutation as that detected in blood. Although the PMA yields accurate and reproducible results over a broad range of mutant and
E.F. Bowen et al./Journal of Virological Methods" 68 (1997) 225 234
w i l d - t y p e p r o p o r t i o n s , the assay is less reliable in p r o v i d i n g a c c u r a t e d a t a b e l o w 5% o f either genotype. T h e d a t a o b t a i n e d at c o d o n 595 for p a t i e n t s 1 a n d 2 suggest t h a t a low level prevalence o f m u t a n t U L 9 7 virus m a y be p r e s e n t in the H C M V p o p u l a t i o n in the absence o f ganciclovir. I n the case o f herpes simplex virus, T K m i n u s g e n o t y p e s are p r e s e n t at low levels in the absence o f aciclovir ( a p p r o x i m a t e l y 1 in 10 0 0 0 ) w h i c h can s u b s e q u e n t l y be given a g r o w t h a d v a n t a g e b y t r e a t m e n t with aciclovir (reviewed b y Collins a n d D a r b y , 1991). F u r t h e r w o r k is r e q u i r e d to ascertain the frequency with which U L 9 7 resistant m u t a n t s are p r e s e n t within the h u m a n h o s t a n d the influence o f this frequency o n the time to resurgence o f v i r a e m i a a n d emergence o f resistant virus. I n conclusion, the d e v e l o p m e n t a n d a p p l i c a t i o n o f a r a p i d p o i n t m u t a t i o n assay to assess the q u a n t i t a t i v e changes in the d i s t r i b u t i o n o f U L 9 7 m u t a n t s c o n f e r r i n g resistance to G C V in vivo have been described. T h e a p p l i c a t i o n o f this m e t h o d to the r o u t i n e m o n i t o r i n g o f p a t i e n t s e x p o s e d to G C V will allow the emergence o f G C V resistant strains o f H C M V to be assessed a n d a l t e r a t i o n s in t h e r a p y considered. In a d d i t i o n , it s h o u l d facilitate the investigation o f the relative fitness o f m u t a n t a n d w i l d - t y p e U L 9 7 strains o f H C M V as has been u n d e r t a k e n for H I V ( G o u d s m i t et al., 1996).
Acknowledgements This w o r k was s u p p o r t e d b y the U K M e d i c a l R e s e a r c h C o u n c i l t h r o u g h an M R C t r a i n i n g fellowship to D r E.F. B o w e n a n d the W e l l c o m e Trust. W e are grateful to the staff a n d p a t i e n t s o f the I a n C h a r l e s o n D a y Centre for their p a r t i c i p a t i o n in this study. W e are grateful to M r Steve K a y e for helpful discussions.
References Alain, S., Mazeron, M.C., Pepin, J.M., Bergmann, J.F., Narwa, R., Raskine, L., Sanson-Le Pors, M.J., 1995. Value of a new rapid non-radioactive sequencing method for analysis of the cytomegalovirus UL97 gene in ganciclovirresistant strains. J. Virol. Methods 51 (2-3), 241 251. Baldanti, F., Underwood, M.R., Stanat, S.C., Biron, K.K.,
233
Chou, S., Sarasini, A., Silini, E., Gerna, G., 1996. Single amino acid changes in the DNA polymerase confer foscarnet resistance and slow-growth phenotype, while mutations in the UL97-encoded phosphotransferase confer ganciclovir resistance in three double-resistant human cytomegalovirus strains recovered from patients with AIDS. J. Virol. 70 (3), 1390-1395. Boivin, G., Chou, S., Quirk, M.R., Erice, A., Jordan, M.C., 1996. Detection of ganciclovir resistance mutations quantitation of cytomegalovirus (CMV) DNA in leukocytes of patients with fatal disseminated CMV disease. J. Infect. Dis. 173 (3), 523 528. Bowen, E.F., Wilson, P., Cope, A., Sabin, C., Griffiths, P., Davey, C., Johnson, M., Emery, V., 1996. Cytomegalovirus retinitis in AIDS patients: influence of cytomegaloviral load on response to ganciclovir, time to recurrence and survival. AIDS 10 (13), 1515 1520. Chee, M.S., Bankier, A.T., Beck, S., Bohni, R., Brown, C.M., Cerny, R., Horsnell, T., Hutchison, C.A. 3rd, Kouzarides, T., Martignetti, J.A., et al., 1990. Analysis of the proteincoding content of the sequence of human cytomegalovirus strain AD169. Curr, Top. Microbiol. lmmunol. 154, 125 169. Chou, S., Erice, A., Jordan, M.C., Vercellotti, G.M., Michels, K.R., Talarico, C.L., Stanat, S.C., Birom K.K., 1995a. Analysis of the UL97 phosphotransferase coding sequence in clinical cytomegalovirus isolates and identification of mutations conferring ganciclovir resistance. J. Infect. Dis. 171 (3), 576 583. Chou, S., Guentzel, S., Michels, K.R., Miner, R.C., Drew, W.L., 1995b. Frequency of UL97 phosphotransferase mutations related to ganciclovir resistance in clinical cytomegalovirus isolates. J. Infect. Dis. 172 (1), 239-242. Collins, P., Darby, G., 1991. Laboratory studies of herpes simplex virus strains resistant to acyclovir. Rev. Med. Virol. 1, 19 28. Crumpacker, C.S., 1996. Ganciclovir. New Engl. J. Med 335 (10), 721-729. Darlington, J., Super, M., Patel, K., Grundy, J.E., Griffiths, P.D., Emery, V.C., 1991. Use of the polymerase chain reaction to analyse sequence variation within a major neutralizing epitope of glycoprotein B (gp58) in clinical isolates of human cytomegalovirus. J. Gen. Virol. 72 (8), 1985 1989. Drew, W.L., Miner, R.C., Busch, D.F., Follansbee, S.E., Gullett, J., Mehalko, S.G., Gordon, S.M., Owen, W.F. Jr., Matthews, T.R., Buhles, W.C., et al., 1991. Prevalence of resistance in patients receiving ganciclovir for serious cytomegalovirus infection. J. Infect. Dis. 163 (4), 716 719. Fox, J.C., Griffiths, P.D., Emery, V.C., 1992. Quantification of human cytomegalovirus DNA using the polymerase chain reaction. J. Gen. Virol. 73 (9), 2405 2408. Fox, J.C., Kidd, I.M., Griffiths, P.D., Sweny, P., Emery, V.C., 1995. Longitudinal analysis of cytomegalovirus load in renal transplant recipients using a quantitative polymerase chain reaction: correlation with disease. J. Gen. Virol. 76 (2), 309-319.
234
E.F. Bowen et al. /Journal of Virological Methods' 68 (1997) 225-234
Goudsmit, J., De Ronde, A., Ho, D.D., Perelson, A.S., 1996. Human immunodeficiency virus fitness in vivo: calculations based on a single zidovudine resistance mutation at codon 215 of reverse transcriptase. J. Virol. 70 (8), 5662-5664. Hanson, M.N., Preheim, L.C., Chou, S., Talarico, C.L., Biron, K.K., Erice, A., 1995. Novel mutation in the UL97 gene of a clinical cytomegalovirus strain conferring resistance to ganciclovir. Antimicrob. Agents Chernother. 39 (5), 12041205. Kaye, S., Loveday, C., Tedder, R.S., 1992. A microtitre format point mutation assay: application to the detection of drug resistance in human immunodeficiency virus type-I infected patients treated with zidovudine. J. Med. Virol. 37 (4), 241-246. Lalezari, J.P., Stagg, R.J., Kupperrnann, B.D., Holland, G.N., Kramer, F., Ives, D.V,, Youle, M., Robinson, M.R., Drew, W.L., Jaffe, H.S., 1997. Intravenous cidofovir for peripheral cytomegalovirus retinitis in patients with AIDS. A randomized, controlled trial. Ann. Intern. Med. 126 (4), 257-263. Littler, E., Stuart, A.D., Chee, M.S., 1992. Human cytomegalovirns UL97 open reading frame encodes a protein that phosphorylates the antiviral nucleoside analogue ganciclovir. Nature 358 (6382), 160-162. Lurain, N.S., Spafford, L.E., Thompson, K.D., 1994. Muta-
tion in the UL97 open reading frame of human cytomegalovirus strains resistant to ganciclovir. J. Virol. 68 (7), 4427 4431. Spector, S.A., McKinley, G.F., Lalezari, J.P., Samo, T., Andruczk, R., Follansbee, S., Sparti, P.D., Havlir, D.V., Simpson, G., et al., 1996. Oral ganciclovir for the prevention of cytomegalovirus disease in persons with AIDS. Roche Cooperative Oral Ganciclovir Study Group. New Engl. J. Med 334 (23), 1491-1497. Studies of Ocular complications of AIDS Research Group (SOCA) in collaboration with the AIDS Clinical Trials Group, 1997. Parenteral cidofovir for cytomegalovirus retinitis in patients with AIDS: the HPMPC peripheral cytomegalovirus retinitis trial. A randomized, controlled trial. Ann. Intern. Med. 126 (4) 264-274. Sullivan, V., Talarieo, C.L., Stanat, S.C., Davis, M., Coen, D.M., Biron, K.K., 1992. A protein kinase homologue controls phosphorylation of ganciclovir in human cytomegalovirus-infected cells. Nature 358 (6382), 162-164. Wolf~ D.G., Smith, I.L., Lee, D.J., Freeman, W.R., FloresAguilar, M., Spector, S.A., 1995. Mutations in human cytomegalovirus UL97 gene confer clinical resistance to ganciclovir and can be detected directly in patient plasma. J. Clin. Invest. 95 (1), 257-263.
ELSEVIER
Journal of Virological Methods 68 (1997) 225 234
Development of a point mutation assay for the detection of human cytomegalovirus UL97 mutations associated with ganciclovir resistance E.
Frances Bowen a, Margaret A. Johnson b, Paul D. Griffiths a, Vincent C. Emery
a,,
a Department of Virology Royal Free Hospital School of Medieine, Rowland Hill Street, London N W 3 2QG, UK b AIDS Unit, Royal Free Hospital School of Medicine, Rowland Hill Street, London N W 3 2QG, UK
Received 26 March 1997: received in revised form 29 July 1997; accepted 29 July 1997
Abstract A point mutation assay was developed to detect the quantitative prevalence of mutations at codons 460 (M to I; M to V), 520 (H to Q), 594 (A to V) and 595 (L to F: L to S) within the UL97 gene of human cytomegalovirus which segregate with ganciclovir resistance. Synthetic mixtures of wild-type and mutant plasmids containing the UL97 gene were amplified by nested polymerase chain reaction and the 700 base pair amplicon subsequently subjected to the point mutation assay. In plasmid reconstruction experiments, there was a high correlation between experimentally derived percentage mutant with the theoretical values. The assay was then used to assess the changes in the genetic composition of the UL97 gene in three patients on prolonged ganciclovir therapy. All three patients developed genotypic resistance against ganciclovir involving mutation at codon L595S, L595F and double mutation at codons L595F and M460I. In one patient, alteration of therapy to foscarnet did not affect the composition of UL97 and virus remained genotypically resistant to ganciclovir. In contrast, in two patients whose therapy was altered to cidofovir (HPMPC), repopulation with cytomegalovirus strains carrying the wild-type (ganciclovir-sensitive) codon at positions 595 and 460 occurred. The potential use of this assay for the rapid detection of cytomegalovirus resistance in patients on long-term ganciclovir therapy is discussed. © 1997 Elsevier Science B.V. Keywords: Cidofovir; Resistance; Therapy; Dynamics
1. Introduction
* Corresponding author.
Since 1989, ganciclovir (GCV) has been the first line treatment for h u m a n cytomegalovirus ( H C M V ) disease in patients with Acquired Ira-
0166-0934/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PH S0166-0934(97)00131-6
226
E.F. Bowen et al. /Journal of Virological Methods' 68 (1997) 225 234
munodeficiency Syndrome (AIDS) (reviewed by Crumpacker, 1996). This widespread usage has led to the emergence of GCV resistant strains of HCMV that are becoming more problematic as patients with HCMV retinitis are living longer and hence receiving longer courses of ganciclovir. In addition, with the results of the oral ganciclovir study 1654 there is likely to be increasing use of prophylactic GCV (Spector et al., 1996). The frequency of resistant strains has been estimated at 7.6% of patients receiving GCV for more than 3 months (Drew et al., 1991). However, it should be noted that these resistant strains were identified in urine samples after passage in vitro and may not truly represent resistance patterns in vivo. In 1992, two elegant molecular studies demonstrated that the HCMV UL97 open reading frame was responsible for the monophosphorylation of ganciclovir (Littler et al., 1992; Sullivan et al., 1992). This discovery has allowed the development of molecular techniques such as PCR to amplify UL97 both from cell culture propagated strains of HCMV derived from patients or directly from clinical samples. The UL97 DNA sequence in clinical samples (predominately plasma) has been determined by direct sequencing methods using 33p end-labelled primers, the incorporation of 35Slabelled dATP (Chou et al., 1995a; Hanson et al., 1995; Wolf et al., 1995) or silver staining of the sequencing products (Alain et al., 1995). From these studies, the most frequently identified mutations in clinical samples obtained from patients on prolonged GCV therapy are at amino acids M460V, M460I, H520Q, A594V, L595S and L595F (Lurain et al., 1994; Alain et al., 1995; Chou et al., 1995b; Hanson et al., 1995; Wolf et al., 1995; Baldanti et al., 1996; Boivin et al., 1996). All of these mutations have been shown by marker transfer studies to confer GCV resistance to Ad169. Five of these mutations (V460, Q520, V594, $595 and F595) can be detected by a combination of PCR followed by restriction endonuclease digestion (Chou et al., 1995b; Hanson et al., 1995). Although these methods potentially allow the rapid identification of resistant virus in clinical samples, they do not provide quantitative information of the prevalence of these mutations within the viral population. Such data will be
important in monitoring the evolution of GCV resistance of HCMV in vivo and to allow modifications to therapy at appropriate times. In order to gain insight into the quantitative prevalence of GCV resistance genotypes in patients on prolonged GCV therapy, a rapid point mutation assay (PMA) was developed to screen for the mutations M460V, M460I, H520Q, A594V, L595S and L595F and used to determine the quantitative prevalence of UL97 mutations in multiple samples from AIDS retinitis patients receiving prolonged exposure to anti-HCMV therapy.
2. Materials and methods 2.1. Patients
A subset of HIV-seropositive patients who attended the AIDS clinic at the Royal Free Hospital from 1 September 1993 with a diagnosis of cytomegalovirus retinitis were used to assess the performance of the point mutation assay (Darlington et al., 1991; Bowen et al., 1996). Samples of blood (4 ml in citrate filled tubes) and urine were collected before treatment (on the day of diagnosis of retinitis), at the end of the induction period, and then monthly thereafter. Extra samples of blood and urine were collected at times of retinitis progression. All patients received ganciclovir for induction therapy and as primary maintenance therapy either intravenously or orally. Intolerance to ganciclovir or recurrent progression of HCMV retinitis was treated with either foscarnet or cidofovir. Three patients in whom substantial numbers of blood samples ( > 10) were available were selected for detailed analysis with the point mutation assay described below. 2.2. Polymerase chain reaction
Initially all samples were subjected to a qualitative PCR for the HCMV gene gB (Darlington et al., 1991; Bowen et al., 1996). All HCMV positive samples were then quantitated using a quantitative-competitive PCR previously described (Fox et
E.F. Bowen et al./Journal of Virological Methods 68 (7997) 225 234 al., 1992, 1995; Bowen et al., 1996) and analysed for UL97. A nested PCR for the detection of UL97 was developed using D N A extracted from whole blood using a commercial ion-exchange column (Qiagen, UK). The outer primers were as follows: 1. UL97 A 5' A G A C G G T G C T A C G G T C T G G A T G T 3' 2. UL97 B 5' G T T T G T A C C T T C T C T G T T G C C T T T 3' Each PCR reaction contained 150 ng primer, 3 m M MgC12, 200 # M of each dNTP and 2 units of Taq polymerase in 10 x PCR buffer containing 250 m M Tris-HC1 pH 8.4, 170 m M ammonium sulphate, 100 m M fl-mercaptoethanol, 0.02% (w/ v) gelatin. Following the addition of 5 /~1 target D N A the reaction was made up to a final volume of 50 ill with SDW. Cycling conditions were as follows; 95°C for 6 min followed by 25 cycles of 94°C 1 min, 45°C 2 min, 72°C 3 min; 2 tL1 of the outer PCR product was then amplified in a nested PCR. In each nested PCR reaction one of the primers was 5' biotinylated to enable adherence of one D N A strand to the wells of a streptavidincoated microtitre plate. The nested primer sequences were as follows: 1. UL97 C 5' C A A C G T C A C G G T A C A T C G A C GTTT 2. UL97 D 5' G C C A T G C T C G C C C A G G A G A CAGG In each PCR, 50 ng of biotinylated primer (either primer C or D depending upon the specific P M A probe used; see later) was used in conjunction with 100 ng non-biotinylated primer. The inner round PCR conditions were the same as described for the outer-round PCR except for the use of 1.5 m M MgC12 in a total reaction volume of 100 itl. Cycling conditions were 35 cycles of 94°C 1 rain, 58°C 2 min, 72°C 3 rain. The presence of the 700 bp amplicons were detected by electrophoresing 10 /~1 PCR product on a 1% agarose gel in the presence of a 2 HindIII/EcoR1 D N A marker. 2.3. Design of oligonucleotide probes The oligonucleotide probes were designed to bind to the biotinylated strand of the PCR
227
product and terminated one base prior to the point mutation to be detected. The sequences for the probes were taken from the published sequence of either the sense or anti-sense strand of Ad169 (Chee et al., 1990). In areas of the UL97 gene where common silent polymorphisms were known to occur in clinical samples a mixed base was included in the probe sequence. All oligonucleotide probes were synthesised and H P L C purified by Cruachem, UK. Probes were made to detect the six most common mutations; M460V (ACT-to-GTG), M460I (ATG-to-ATT), H520Q (CAC-to-CAG), A594V (GCG-to-GTG), L595S (TTG-to-TCG) and L595F (TTG-to-TTT). Probes I460, V594, $595 and F595 were complementary to the biotinylated sense strand of the PCR product whilst probes V460 and Q520 bound to the biotinylated anti-sense strand. The probe sequences were as follows: 1. Probe 1: L595F; 5' A G T G C G T G A G C T T A / G C C G T T C T C antisense ~ 5'B UL97C 2. Probe 2: L595S; 5' G T G C G T G A G C T T A / G C C G T T C T C C antisense --* 5'B UL97C 3. Probe 3: A594V; 5 ' G T G A G C T T A / G C C G T T CTCCAAC antisense --, 5'B UL97C 4. Probe 4: H520Q; 5' C C G T C T G C G C G A A T G T T A C C A sense -~ 5'B UL97D 5. Probe 5: M460I; 5' G T T C A C G T C G A T G A G C A C G T T antisense --, 5'BUL97C 6. Probe 6: M460V; 5' T G C C A C T T T G A C / T A T T A C A C C C sense ~ 5'B UL97D where 5'B indicates the presence of biotin on either primer C or D. 2.4. Microtitre point mutation assay The microtitre point mutation assay (PMA) to detect H C M V UL97 mutations was based on the approach originally described by Kaye et al. (1992) for the detection of zidovudine specific mutations in the R T gene of HIV-1. Minor modifications to optimise the performance of this assay for UL97 were effected. These included the use of 96-well streptavidin-coated plates produced by Advanced Biotechnologies Ltd., optimisation of the binding of the biotinylated PCR product to the plate and the use of a lower annealing temperature during the hybridisation of the probe to the biotinylated strand.
228
E.F. Bowen et al./Journal of Virological Methods 68 (J997) 225 234
Following the nested PCR, 10 /~1 of biotinylated PCR product were diluted to 25/~1 with 1 x T T A buffer (10 x TTA; 100 m M Tris-HC1 pH 7.6, 0.5% Tween 20, 1.0% sodium azide stored at room temperature) and added to each of four microtitre wells of a 96-well streptavidin-coated plate (400 ng affinity purified streptavidin/300 Itl well; StrepMax, Advanced Biotechnologies Ltd.) and incubated at 37°C for 60 rain. The wells were washed three times with TTA buffer and then 40 ¢tl 0.15 M N a O H were added to each well and left to incubate for 5 rain at room temperature to denature the captured PCR product. The wells were washed four times with T T A to remove the released second strand and 25/~1 of annealing mix made up of the appropriate oligonucleotide probe in 2.5 ml P M A diluent (PMA diluent; 40 mM Tris HC1 pH 7.6, 20 mM MgC12, 50 m M NaC1 stored at 4°C) was added to each well. To optimise the PMA for UL97, oligonucleotide probe concentrations from 3.3 ng to 26 ng were used to maximise probe binding to the biotinylated product whilst keeping non-specific binding to a minimum. The plate was covered and placed in a 55°C water bath for 3 rain and then allowed to cool slowly to room temperature over at least 30 min; 24/~1 of each of the four 35S-labelled dNTPs (1200 Ci/mmol (NEN DuPont) stored at - 2 0 ° C as a 1 in 10 dilution in PMA diluent) was added to a 168-/~1 aliquot of 0.1 M DTT. To this mixture was then added Klenow polymerase (10 units, Boehringer Mannheim), diluted in 200 /~1 P M A diluent. A 9.8-~1 aliquot of the resultant labelling mix was added to each well with a different dNTP being added to each of the four wells, and incubated at room temperature for 3 min. The wells were washed four times with TTA to remove unincorporated label from the target strand and 40 i~1 of 0.15 M N a O H was then added to denature the probe and any incorporated labelled nucleotide. The N a O H solution was removed and mixed with 5 ml scintillation cocktail (Optiphase Hisafe-3) in 10-ml scintillation vials (Camlab, Cambridge, UK). The radioactivity in the samples was determined in a LKB Rackbeta scintillation counter.
2.5. Generation o f the standard curve f o r mixtures o f wild-type and mutant codons
In order to establish a standard curve for the differentiation of mutant and wild-type sequences by the PMA, standard amounts of UL97 D N A of a known sequence were amplified by PCR and subjected to the PMA. For these experiments three separate UL97 clones derived by PCR amplification of blood from patients on long-term GCV therapy (which had been previously analysed by conventional D N A sequencing methods of cloned UL97 amplicons) were used. Each clone possessed a single mutation at codons 594 (A to V), 595 (L to S) and 595 (L to F). In each PCR the ratio of the two genotypes was mixed as follows; 100% WT, 75% WT: 25% MUT, 50% WT: 50% MUT, 25% WT: 75% MUT, 100% MUT. Each clone was used as either WT or M U T depending on the probe used. For example, when mixing A594V and L595S, probe 3 detects A595V as M U T and L595S as WT, whilst probe 2 detects L595S as M U T and A595V as WT. Each reaction was performed in triplicate. Probes 4 - 6 were standardised on clones that were known to contain wild-type sequences at that particular mutation. Linear regression fitting was used to determine the adherence of the data points to a straight line.
3. Results 3. I. Generation o f the standard curves
The ability of the optimised point mutation assay described in the methods section to detect different ratios of wild-type or mutant nucleotides at specific codons was assessed by analysis of pre-determined mixtures obtained from previously cloned and sequenced UL97 products derived from clinical samples. The data from three independent experiments were used to construct standard curves for L595F, L595S and A595V relating predicted ratios to wild-type: mutant to the experimentally determined ratios. The results are shown in Fig. 1 together with the linear regression line-fit. In each PMA the experimentally deter-
E.F. Bowen et al./Journal of Virological Methods 68 (1997) 225 234 100 -
100 -
80
80
so
_~ 6O ._o
~ 40
u_ 40 U3
229
03
20 P = O" 0002
0 2
0
'
I
20
'
I
'
40
I
'
I
60
'
20
--1-
0 :I 0
I
80
-
100
'
I
P
'
20
I
40
% 595 S (theoretical)
'
I
'
60
I
80
=0"0008
'
I
100
% 595 F (theoret!cal)
100 -
80
_~ 60 o > 40 O'1 Lf)
~0.~/
P = 0.0001
z/ '
0
I
20
'
l
40
'
I
J
60
l
80
'
I
100
% 594 V (theoretical)
Fig. 1. Calibration curves relating the experimentally determined ratio of wild-type and mutant genotypes at codons L595F, L595S and M460V to the predicted ratio. Each data point was generated from three separate PMAs and the mean + 1 standard deviation are quoted on the graph. The line of best fit was computed using the line fitting program available with Microcal Origin. The adherence of line to the data is given by the R-value and its associated significance indicated. m i n e d r a t i o o f w i l d - t y p e to m u t a n t was in close a g r e e m e n t to the p r e d i c t e d ratios.
3.2. Analysis o f clinical samples Jor the presence o f UL97 mutations I n o r d e r to ascertain w h e t h e r the P M A c o u l d be used successfully on clinical samples, m u l t i p l e sam-
ples were a n a l y s e d f r o m three p a t i e n t s w h o develop e d H C M V retinitis a n d were initially t r e a t e d with ganciclovir t h e r a p y , then with either f o s c a r n e t o r c i d o f o v i r due the a p p a r e n t clinical failure o f ganciclovir. It was r e a s o n e d t h a t these p a t i e n t s were likely to have ganciclovir resistant strains o f H C M V on the basis o f their clinical h i s t o r y a n d so w o u l d p r o vide suitable samples on which to assess the assay.
230
E.F. Bowen et al./Journal of Virological Methods 68 (1997) 225-234
Each blood DNA extract was subjected to the PMA for the five most common coding mutations that have been shown to endow ganciclovir resistance to Ad169 (see above). In all cases, pre-treatment samples were also analysed by the PMA. The results are shown in Table 1 and indicate the percentage of mutant virus present at codons 595, 594, 520 and 460 for each patient at multiple time-points. The anti-HCMV therapy in use at the time the sample was obtained is also included. In patients 1 and 2, the PMA identified the presence of a low but reproducible amount of the L595S and L595F mutations respectively at the diagnosis of HCMV retinitis, i.e. before either patient had been exposed to ganciclovir. In contrast the PMA results for patient 3 indicated only wild-type sequences at all codons. During ganciclovir therapy, the proportion of L595F and L595S genotypes in patients 1 and 2 increased rapidly to reach a level greater than 90% by days 179 and 111 post-diagnosis of retinitis respectively. The substitution of cidofovir for ganciclovir in patient 1 at day 279 resulted in a gradual repopulation with HCMV strains that were wildtype at codon 595. In patient 2, whose therapy was altered to foscarnet at day 272, no substantial alterations in the proportion of L595F genotype were observed. In this patient, HCMV was detected in the urine at day 308 and was predominately wild-type at codon 595 (95%) despite the presence of predominately mutant virus in the blood. However, mutant virus at this codon rapidly repopulated the virus strains and within 3 weeks of GCV therapy for gastrointestinal disease was approximately 90% mutant at codon 595. The genotypic resistance profile observed in patient 3 is more complex. Both first and second clinical progressions of retinitis were not accompanied by detection of HCMV in the blood by PCR. However, at day 198 virus was detected which was 78% mutant at codon 595 (L to F) and 32% mutant at codon 460 (M to I). Following a short period of foscarnet the patient was initiated on cidofovir therapy. Similar to the results obtained for patient 1, the repopulation of wild-type UL97 genotypes occurred at both the L595F and M460I codons. Viral load in all three patients remained elevated following the development of genotypic resistance to ganciclovir in UL97 despite the intervention with
drugs that do not require UL97 for activation. A ty-pical graph of the viral load profile and alterations in the genotypic composition of UL97 is shown in Fig. 2.
4. Discussion
The development of a point mutation assay is described which provides a rapid and reproducible means of analysing genetic changes in UL97 amplified directly from clinical samples. The method allows the prospective study of changes in UL97 under the influence of different anti-HCMV drugs without the need for in vitro passage. To our knowledge this is the first time that quantitative prevalence of mutations in patients undergoing antiHCMV therapy has been detailed. The assay was similar with respect to reproducibility and ease of use to that previously described for HIV RT mutations segregating with AZT resistance (Kaye et al., 1992). Prospective analysis of multiple samples derived from three patients treated with GCV for HCMV retinitis using the PMA revealed the presence of one or more mutations in codons 595 and 460 of the UL97 gene which are known from other studies to confer reduced susceptibility to GCV. In these patients the availability of multiple PCR positive follow-up samples enabled the evolution of GCV resistance at the UL97 locus to be examined and the influence of alternations in therapy on the re-emergence of UL97 wild-type strains investigated. In patient 1, the virus was > 90% mutant at codon 595 at the time of resurgence of PCR viraemia (day 179) and the first clinical progression of the retinitis was observed at day 241. Due to the failure of this individual to respond to GCV re-induction therapy and evidence of further clinical progression the patient was switched to cidofovir therapy at day 279. Over the ensuing 151 days outgrowth of the wild-type UL97 (L595) virus occurred. This observation is consistent with the fact that cidofovir inhibits CMV replication through the DNA polymerase and hence UL97 mutants have no selective growth advantage under the new circumstances. Despite the antiviral effects of cidofovir, the viral load in this patient remained
E.F. Bowen et al. /Journal of Virological Methods 68 (1997) 225-234
231
Table 1 Distribution of genotypes at codons L595F, L595S, A594V, H520Q, M460I and M460V in the UL97 gene of H C M V in multiple blood samples from three patients by the point mutation assay Pt
HCMV Diagnosis
Days
Drug
1
Retinitis
0 179 241 248 274 279 301 329 361 375 403 430
-GCV GCV GCV GCV CDV CDV CDV CDV CDV CDV CDV
wt wt wt wt wt wt wt wt wt wt wt wt
0 93 111 127 134 180 204 225 272 308 322 397 438 468
GCV GCV GCV GCV GCV GCV GCV PFA PFA PFA PFA PFA PFA
9% 79% 95% 99% 99.1% 97% 99.6% 97% 98% 97% 97% 97% 95% 88%
0 77 133 198 230 247 280 296 301 320 330
GCV GCV GCV GCV PFA CDV CDV CDV CDV CDV
wt -ve - ve 78% 79% 77% 81% 78% 66% 24% 2%
1st R P 2nd R P
2
Retinitis 1st R P
2nd R P 3rd R P
3
Retinitis 1st R P 2nd R P
3rd R P
L595F (%F)
L595S (%S)
8%
4% - ve - ve - ve -re - ve -ve - ve - ve 5% N/A 87% 90%, 92%
A594V (%V)
H520Q (%Q)
M460I (%I)
M460V (%V)
94% 97% 96% 95% 90.5% 70% 61% 19%, 3% 7% 8%
wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt wt wt wt wt wt wt wt wt wt wt wt wt wt
wt -ve - ve wt wt wt wt wt wt wt wt
wt -ve - ve wt wt wt wt wt wt wt wt
wt -ve - ve wt wt wt wt wt wt wt wt
wt -ve - ve 32% 84% 93% 98% 98.5% 98% 22°/,, 4%
wt -ve - ve wt wt wt wt wt wt wt wt
In patient 2, the % m u t a n t at codon L595F is quoted for blood (left-hand data set) and urine (right-hand data set). R P = retinitis progression; G C V = ganciclovir, P F A = foscarnet; C D V = cidofovir, wt = wild-type at codon of interest. high
implying
that
resistant to GCV polymerase
CMV
strains
may
be dually
and cidofovir at the UL54 (DNA
l o c u s ) w h e r e p r e s e n t . T h i s p o s s i b i l i t y is
currently
being investigated.
observed
in patient
evolution of HCMV
A similar profile was
3, e x c e p t
that
in this
case
strains carrying two mutations
in UL97 were observed
(codons
595 and 460), both
of which
reverted
to wild-type
of cidofovir. Evolution was
not
between
parallel
following
exposure
of resistance at each codon
implying
differences
viral strains carrying
in
fitness
single mutations
codons 595 and 460 or dual mutations. of cidofovir to inhibit systemic HCMV
at
The inability replication
in spite of its effects at stabilising HCMV
retinitis
E.F. Bowen et al. /Journal of Virological Methods 68 (1997) 225-234
232 HCMV
RETINITIS
RP
RP
RP
--
100
80
60 % 595 S or 4601
HCMV log10 genomes/ml blood
40
20
0 ,
0
100
200
300
I
400
500
DAYS
GCV Induction
RP
RETINITIS PROGRESSION
GCV Maintenance
[]
HCMV load
Intravitreal FOSCARNET
•
% 595 S
CIDOFOVIR
O
% 460 I
Fig. 2. Longitudinal viral load profiles and alterations of UL97 genotypic resistance in patient 3 of Table 1. Periods and type of therapy are indicated together with clinical progression of HCMV retinitis.
has been recently reported (Lalezari et al., 1997; SOCA, 1997). In patient 2, evolution of high level GCV resistance at UL97 was observed over a period of 111 days which did not alter following exposure of the patient to foscarnet. Since foscarnet also targets the UL54 gene product of HCMV directly, it is intriguing why the viral species was not repopulated by strains carrying the wild-type codon at position 595. Since viral load remained elevated in this patient during foscarnet therapy it is plausible that this individual harboured DNA polymerase mutants resistant to GCV and foscarnet. Again this
possibility is currently under investigation. The temporal dislocation of virus detection in different body sites is aptly demonstrated in this patient. Despite the presence of GCV resistance virus in blood at high viral loads for over 200 days, virus was detected in urine relatively late in the clinical course. Initially, this virus was wild-type at codon 595 but the population was rapidly re-populated by a strain carrying the same GCV resistant mutation as that detected in blood. Although the PMA yields accurate and reproducible results over a broad range of mutant and
E.F. Bowen et al./Journal of Virological Methods" 68 (1997) 225 234
w i l d - t y p e p r o p o r t i o n s , the assay is less reliable in p r o v i d i n g a c c u r a t e d a t a b e l o w 5% o f either genotype. T h e d a t a o b t a i n e d at c o d o n 595 for p a t i e n t s 1 a n d 2 suggest t h a t a low level prevalence o f m u t a n t U L 9 7 virus m a y be p r e s e n t in the H C M V p o p u l a t i o n in the absence o f ganciclovir. I n the case o f herpes simplex virus, T K m i n u s g e n o t y p e s are p r e s e n t at low levels in the absence o f aciclovir ( a p p r o x i m a t e l y 1 in 10 0 0 0 ) w h i c h can s u b s e q u e n t l y be given a g r o w t h a d v a n t a g e b y t r e a t m e n t with aciclovir (reviewed b y Collins a n d D a r b y , 1991). F u r t h e r w o r k is r e q u i r e d to ascertain the frequency with which U L 9 7 resistant m u t a n t s are p r e s e n t within the h u m a n h o s t a n d the influence o f this frequency o n the time to resurgence o f v i r a e m i a a n d emergence o f resistant virus. I n conclusion, the d e v e l o p m e n t a n d a p p l i c a t i o n o f a r a p i d p o i n t m u t a t i o n assay to assess the q u a n t i t a t i v e changes in the d i s t r i b u t i o n o f U L 9 7 m u t a n t s c o n f e r r i n g resistance to G C V in vivo have been described. T h e a p p l i c a t i o n o f this m e t h o d to the r o u t i n e m o n i t o r i n g o f p a t i e n t s e x p o s e d to G C V will allow the emergence o f G C V resistant strains o f H C M V to be assessed a n d a l t e r a t i o n s in t h e r a p y considered. In a d d i t i o n , it s h o u l d facilitate the investigation o f the relative fitness o f m u t a n t a n d w i l d - t y p e U L 9 7 strains o f H C M V as has been u n d e r t a k e n for H I V ( G o u d s m i t et al., 1996).
Acknowledgements This w o r k was s u p p o r t e d b y the U K M e d i c a l R e s e a r c h C o u n c i l t h r o u g h an M R C t r a i n i n g fellowship to D r E.F. B o w e n a n d the W e l l c o m e Trust. W e are grateful to the staff a n d p a t i e n t s o f the I a n C h a r l e s o n D a y Centre for their p a r t i c i p a t i o n in this study. W e are grateful to M r Steve K a y e for helpful discussions.
References Alain, S., Mazeron, M.C., Pepin, J.M., Bergmann, J.F., Narwa, R., Raskine, L., Sanson-Le Pors, M.J., 1995. Value of a new rapid non-radioactive sequencing method for analysis of the cytomegalovirus UL97 gene in ganciclovirresistant strains. J. Virol. Methods 51 (2-3), 241 251. Baldanti, F., Underwood, M.R., Stanat, S.C., Biron, K.K.,
233
Chou, S., Sarasini, A., Silini, E., Gerna, G., 1996. Single amino acid changes in the DNA polymerase confer foscarnet resistance and slow-growth phenotype, while mutations in the UL97-encoded phosphotransferase confer ganciclovir resistance in three double-resistant human cytomegalovirus strains recovered from patients with AIDS. J. Virol. 70 (3), 1390-1395. Boivin, G., Chou, S., Quirk, M.R., Erice, A., Jordan, M.C., 1996. Detection of ganciclovir resistance mutations quantitation of cytomegalovirus (CMV) DNA in leukocytes of patients with fatal disseminated CMV disease. J. Infect. Dis. 173 (3), 523 528. Bowen, E.F., Wilson, P., Cope, A., Sabin, C., Griffiths, P., Davey, C., Johnson, M., Emery, V., 1996. Cytomegalovirus retinitis in AIDS patients: influence of cytomegaloviral load on response to ganciclovir, time to recurrence and survival. AIDS 10 (13), 1515 1520. Chee, M.S., Bankier, A.T., Beck, S., Bohni, R., Brown, C.M., Cerny, R., Horsnell, T., Hutchison, C.A. 3rd, Kouzarides, T., Martignetti, J.A., et al., 1990. Analysis of the proteincoding content of the sequence of human cytomegalovirus strain AD169. Curr, Top. Microbiol. lmmunol. 154, 125 169. Chou, S., Erice, A., Jordan, M.C., Vercellotti, G.M., Michels, K.R., Talarico, C.L., Stanat, S.C., Birom K.K., 1995a. Analysis of the UL97 phosphotransferase coding sequence in clinical cytomegalovirus isolates and identification of mutations conferring ganciclovir resistance. J. Infect. Dis. 171 (3), 576 583. Chou, S., Guentzel, S., Michels, K.R., Miner, R.C., Drew, W.L., 1995b. Frequency of UL97 phosphotransferase mutations related to ganciclovir resistance in clinical cytomegalovirus isolates. J. Infect. Dis. 172 (1), 239-242. Collins, P., Darby, G., 1991. Laboratory studies of herpes simplex virus strains resistant to acyclovir. Rev. Med. Virol. 1, 19 28. Crumpacker, C.S., 1996. Ganciclovir. New Engl. J. Med 335 (10), 721-729. Darlington, J., Super, M., Patel, K., Grundy, J.E., Griffiths, P.D., Emery, V.C., 1991. Use of the polymerase chain reaction to analyse sequence variation within a major neutralizing epitope of glycoprotein B (gp58) in clinical isolates of human cytomegalovirus. J. Gen. Virol. 72 (8), 1985 1989. Drew, W.L., Miner, R.C., Busch, D.F., Follansbee, S.E., Gullett, J., Mehalko, S.G., Gordon, S.M., Owen, W.F. Jr., Matthews, T.R., Buhles, W.C., et al., 1991. Prevalence of resistance in patients receiving ganciclovir for serious cytomegalovirus infection. J. Infect. Dis. 163 (4), 716 719. Fox, J.C., Griffiths, P.D., Emery, V.C., 1992. Quantification of human cytomegalovirus DNA using the polymerase chain reaction. J. Gen. Virol. 73 (9), 2405 2408. Fox, J.C., Kidd, I.M., Griffiths, P.D., Sweny, P., Emery, V.C., 1995. Longitudinal analysis of cytomegalovirus load in renal transplant recipients using a quantitative polymerase chain reaction: correlation with disease. J. Gen. Virol. 76 (2), 309-319.
234
E.F. Bowen et al. /Journal of Virological Methods' 68 (1997) 225-234
Goudsmit, J., De Ronde, A., Ho, D.D., Perelson, A.S., 1996. Human immunodeficiency virus fitness in vivo: calculations based on a single zidovudine resistance mutation at codon 215 of reverse transcriptase. J. Virol. 70 (8), 5662-5664. Hanson, M.N., Preheim, L.C., Chou, S., Talarico, C.L., Biron, K.K., Erice, A., 1995. Novel mutation in the UL97 gene of a clinical cytomegalovirus strain conferring resistance to ganciclovir. Antimicrob. Agents Chernother. 39 (5), 12041205. Kaye, S., Loveday, C., Tedder, R.S., 1992. A microtitre format point mutation assay: application to the detection of drug resistance in human immunodeficiency virus type-I infected patients treated with zidovudine. J. Med. Virol. 37 (4), 241-246. Lalezari, J.P., Stagg, R.J., Kupperrnann, B.D., Holland, G.N., Kramer, F., Ives, D.V,, Youle, M., Robinson, M.R., Drew, W.L., Jaffe, H.S., 1997. Intravenous cidofovir for peripheral cytomegalovirus retinitis in patients with AIDS. A randomized, controlled trial. Ann. Intern. Med. 126 (4), 257-263. Littler, E., Stuart, A.D., Chee, M.S., 1992. Human cytomegalovirns UL97 open reading frame encodes a protein that phosphorylates the antiviral nucleoside analogue ganciclovir. Nature 358 (6382), 160-162. Lurain, N.S., Spafford, L.E., Thompson, K.D., 1994. Muta-
tion in the UL97 open reading frame of human cytomegalovirus strains resistant to ganciclovir. J. Virol. 68 (7), 4427 4431. Spector, S.A., McKinley, G.F., Lalezari, J.P., Samo, T., Andruczk, R., Follansbee, S., Sparti, P.D., Havlir, D.V., Simpson, G., et al., 1996. Oral ganciclovir for the prevention of cytomegalovirus disease in persons with AIDS. Roche Cooperative Oral Ganciclovir Study Group. New Engl. J. Med 334 (23), 1491-1497. Studies of Ocular complications of AIDS Research Group (SOCA) in collaboration with the AIDS Clinical Trials Group, 1997. Parenteral cidofovir for cytomegalovirus retinitis in patients with AIDS: the HPMPC peripheral cytomegalovirus retinitis trial. A randomized, controlled trial. Ann. Intern. Med. 126 (4) 264-274. Sullivan, V., Talarieo, C.L., Stanat, S.C., Davis, M., Coen, D.M., Biron, K.K., 1992. A protein kinase homologue controls phosphorylation of ganciclovir in human cytomegalovirus-infected cells. Nature 358 (6382), 162-164. Wolf~ D.G., Smith, I.L., Lee, D.J., Freeman, W.R., FloresAguilar, M., Spector, S.A., 1995. Mutations in human cytomegalovirus UL97 gene confer clinical resistance to ganciclovir and can be detected directly in patient plasma. J. Clin. Invest. 95 (1), 257-263.