Abstract
purpose. To develop a multiplex polymerase chain reaction (PCR) for the
detection of adenovirus, herpes simplex virus, and Chlamydia
trachomatis in conjunctival swabs.
methods. Oligonucleotide primers for detection of the 3 agents were combined in
one reaction and evaluated for optimal performance using control DNAs
of adenovirus type 2, herpes simplex virus, and C.
trachomatis plasmid. The multiplex PCR was evaluated prospectively
against its corresponding uniplex PCRs, virus isolation, Chlamydia
Amplicor PCR, and an immunoassay technique (immune dot blot test) in a
total of 805 conjunctival swabs from patients with suspected viral and
chlamydial keratoconjunctivitis.
results. The multiplex PCR was as sensitive as uniplex PCRs for the detection of
the agents in clinical specimens. In the prospective study, 48 of 49
(98%) clinical specimens were positive for adenovirus by the multiplex
PCR compared with 26 of 49 (53%) by adenovirus isolation. For herpes
simplex virus detection, the multiplex PCR had a sensitivity of 92%
(34/37) compared with 94.5% (35/37) by cell culture. The multiplex PCR
produced identical results to the Amplicor PCR (21/21; 100%) compared
with 71% (15/21) by the immune dot blot test.
conclusions. With clinical specimens the multiplex PCR was as sensitive as its
respective uniplex PCRs but more sensitive than adenovirus isolation
and as sensitive as herpes simplex virus isolation or C.
trachomatis Amplicor PCR. It has the potential to replace several
diagnostic tests with consequent savings in cost. The test also reduces
the risk of misdiagnosis by the clinicians.
Adenovirus, herpes simplex virus (HSV), and
Chlamydia
trachomatis are common causes of keratoconjunctivitis
(KC).
1 2 3 Outbreaks of adenoviral KC occur throughout the
world.
4 HSV KC occurs in all countries and is the most
common infectious cause of unilateral corneal blindness in the
developed world.
5 In the West, chlamydial KC is caused by
C. trachomatis serovars D to K, and up to 90% of patients
with chlamydial conjunctivitis experience concurrent genital
infection.
6 7 8 9
Several reports have demonstrated the diagnostic advantages of various
laboratory techniques for optimal management of viral and chlamydial
KC,
10 11 12 13 14 15 16 17 18 19 20 but conventional laboratory techniques such as
cell culture and antigen detection methods can be inefficient due to
lack of sensitivity, specificity, and/or speed.
21 During
the last decade, however, several studies have concluded that
polymerase chain reaction (PCR)–based laboratory investigation is a
valuable approach for achieving reliable diagnosis of viral and
chlamydial KC.
21 In this article, we describe a simple and
sensitive multiplex PCR, which detects adenovirus, HSV, and
C.
trachomatis in eye swabs from cases of KC.
A total of 805 eye swabs from 541 patients with suspected viral or
chlamydial KC was tested prospectively for the presence of adenovirus,
HSV, and C. trachomatis within 1 week of receipt. Of the 805
specimens, 456 in virus transport medium (Hanks’ balanced salt
solution, pH 7.4, 10% [vol/vol] fetal calf serum, 2.5% [wt/vol]
sodium bicarbonate, 200 IU/ml penicillin G, 200 mg/ml streptomycin, and
5 mg/ml amphotericin B) had investigation requests for viral etiology,
whereas 349 specimens in sucrose phosphate (2SP) transport medium (8 mM
KH2PO4, 12 mM
K2HPO4, 0.2 M sucrose, 2.5μ
g/ml amphotericin B, 50 μg/ml streptomycin, and 100 μg/ml
vancomycin) had chlamydial investigation requests. All the specimens
were tested for the three agents by the multiplex PCR and its
corresponding uniplex PCRs. The 456 viral specimens were also tested
for the presence of adenovirus and HSV by virus isolation, and the
remaining 349 chlamydial specimens were tested for C.
trachomatis by the Amplicor PCR and Immune Dot Blot (IDB) test.
Cell lines (Vero, HEL 229, and Hep-2) were immediately inoculated
with the untreated eye swabs and maintained for up to 4 weeks to permit
the detection of adenovirus and HSV.
22 23 24
Chlamydial swabs were tested using the Amplicor PCR assay (COBAS
AMPLICOR; Roche Diagnostic Systems, Indianapolis,
IN).
25 This test targets the
C. trachomatis plasmid but uses different primers to those incorporated in our
multiplex PCR.
The IDB test was performed as described previously.
26 Briefly, 0.4 ml of the specimen was digested with 250 μg/ml
proteinase K for 60 minutes at 56°C and then heated to 95°C for 15
minutes. The treated specimen was added to a nitrocellulose membrane in
a dot blot manifold, and the bound chlamydial lipopolysaccharide was
then detected with an
125I-labeled genus-specific
mouse monoclonal antibody.
DNA was extracted from all specimens, including negative material
as extraction negative controls, using lysis buffer.
24 Briefly, equal volumes (75 μl) of sample and lysis buffer (20 mM
Tris–HCl [pH 8.3], 2 mM EDTA, 1% Triton X-100, 0.002% sodium
dodecyl sulfate, and 500 μg/ml proteinase K) were incubated at 56°C
for 2 hours and then boiled for 10 minutes. In addition, a few
specimens were also extracted by guanidinium thiocyanate
(GuSCN)
27 and 30% polyethylene glycol
(PEG).
28
All PCRs were carried out adhering to stringent precautions to
avoid contamination.
28 29 In a 50-μl PCR mixture, all
uniplex PCRs contained 1× PCR buffer (10 mM Tris–HCl [pH 8.3], 50
mM KCl, 0.01% [wt/vol] gelatin), 1.5 mM MgCl
2,
1.25 U amplitaq DNA polymerase, 200 μM of each dNTP, and 0.2 μM of
each primer of the primer pairs ADRJC1/ADRJC2
28 (adenovirus PCR), YS1/YS2
30 (HSV PCR), and
KL1/KL2
31 (
C. trachomatis PCR;
Table 1 ) and 5 μl of appropriate DNA sample or sterile distilled water as a
contamination control. The reaction was overlaid with 2 drops of
mineral oil to prevent evaporation. The assays were performed on a
Programmable Dri-Block PCH-1 (Techne, Cambridge, UK) using one cycle
each of 94°C for 7 minutes, 55°C for 1 minute, and 72°C for 1.5
minute followed by 40 cycles each of 94°C for 1 minute, 55°C for 1
minute, and 72°C for 1.5 minute. The amplification products were
analyzed by electrophoresis in 8% polyacrylamide gels.
28
The reaction conditions of the multiplex PCR were identical to
those described for the uniplex PCRs except for the use of 2.5 U
amplitaq DNA polymerase, 2.2 mM MgCl2, the
presence of equimolar concentrations (0.2 μM) of each primer of the
primer pairs (ADRJC1/ADRJC2, YS1/YS2, and KL1/KL2), and the use of 50
cycles.
Control samples were composed of adenovirus type 2 DNA, HSV type 1
DNA (Life Technologies, Paisley, UK), and the plasmid DNA (pCtL2) of
the
C. trachomatis strain L2/434.
32
The results of the various assays on clinical specimens were
compared using McNemar’s test.
33 P <
0.05 was considered significant.
The development and optimization of both uniplex and multiplex
PCRs required screening for optimal primers and adjustment of PCR
conditions such as enzyme concentration and cycling profile. The
primers ADRJC1/ADRJC2, YS1/YS2, and KL1/KL2 were found satisfactory and
generated PCR products that can be differentiated on gel
electrophoresis
(Table 1) . In addition, the primer pair ADRJC1/ADRJC2
demonstrated great inclusivity for detection of adenovirus serotypes
representing all subgroups found in eyes.
28
In the presence of a single target, the multiplex PCR produced
detection limits of 4 × 102, 3 ×
102, or 102 copies of
adenovirus, HSV, or C. trachomatis plasmid control DNAs,
respectively, whereas detection limits of 103 copies of adenovirus or HSV DNA and 104 copies of C. trachomatis plasmid DNA were achieved when all targets
were present in the same reaction tube. The latter detection limits
were unchanged when only two targets (adenovirus–HSV,
adenovirus–C. trachomatis, or HSV–C.
trachomatis) were added in the mixture. The detection limits of
the uniplex PCRs were 40 copies of adenovirus type 2 DNA, 3 copies of
HSV DNA, and 10 copies of C. trachomatis plasmid DNA.
A total of 805 eye swabs sent to the diagnostic virology
laboratory for virus isolation (456 viral specimens) or
C.
trachomatis detection (349 chlamydia specimens) was tested by the
multiplex PCR and its corresponding uniplex PCRs. The data obtained are
shown in
Table 2 , and an example of the multiplex PCR results is shown in
Figure 1 . The multiplex PCR was compared with virus isolation in the 456 viral
specimens and with Amplicor PCR and IDB test in the 349 chlamydial
specimens.
In all the specimens (n = 805), both the multiplex and
uniplex PCRs produced identical results. Of the viral specimens
(n = 456), 49 specimens were positive for adenovirus
(25 by cell culture and multiplex PCR, 23 by multiplex PCR only, and in
1 sample by cell culture only), and 37 specimens were positive for HSV
(32 were positive by both cell culture and multiplex PCR, 2 specimens
by the multiplex PCR, and 3 were positive by cell culture). In
addition, the multiplex PCR detected C. trachomatis in 10
specimens from patients who had equivalent chlamydia swabs.
In the chlamydial specimens (n = 349), C.
trachomatis was detected in 21 specimens by the multiplex PCR, the
uniplex PCR, and the Amplicor PCR. Of the positive specimens, 15 were
also positive by the IDB test, which was reported negative or equivocal
in the remaining 6 specimens. The multiplex PCR and uniplex PCR
detected viral DNA in 47 chlamydial specimens (45 as adenovirus and 2
as HSV).
If all the positive samples by any of the techniques were considered
true positives, the performance of the uniplex and multiplex PCRs is
identical. For adenovirus detection, the sensitivity of the multiplex
PCR is 98% (48/49), significantly more sensitive than cell culture
(53%; 26/49; P < 0.0001). For HSV detection, the
multiplex PCR has a sensitivity of 92% (34/37), and the corresponding
figure for cell culture is 94.5% (35/37; P = 1.00).
With respect to C. trachomatis detection, the multiplex PCR
and the Amplicor PCR have 100% (21/21) sensitivity, whereas that of
the IDB test is only 71% (15/21; P < 0.05).
All the specimens analyzed (
n = 805) were received from
a total of 541 patients with clinical suspicion of viral and/or
chlamydial KC. Of these patients, viral etiology was suspected in 227
patients, chlamydial etiology in 128 patients, and viral or chlamydial
etiology in the remaining 186 patients
(Table 3) . In the viral etiology group (
n = 227), adenovirus was
detected in 11 patients and HSV in 33 patients. In the group with
clinical suspicion of chlamydial etiology (
n = 128),
C. trachomatis was detected in 8 patients. Interestingly,
adenovirus DNA was detected in 10 patients of this group by the
multiplex and the uniplex PCRs. In the patients with viral or
chlamydial etiology (
n = 186), adenovirus was detected
in 32 patients, HSV in 2 patients, and
C. trachomatis in 10
patients.
Adenovirus, HSV, and
C. trachomatis multiplex PCR
fulfills a rational approach in that all the agents are system (ocular
tissue)- and symptom (KC)-specific and could be of use in all cases of
KC. A reduction in the sensitivity of the multiplex PCR to detect
control DNAs was observed when compared with its corresponding uniplex
PCR. However, in clinical specimens, the multiplex and the uniplex PCRs
produced identical results for the detection of all agents but were
significantly more sensitive than adenovirus isolation, or as sensitive
as HSV isolation or Amplicor PCR for the detection of
C.
trachomatis. The superiority of PCR to adenovirus isolation using
the primer pair ADRJC1/ADRJC2 has been reported by us
previously
28 and is probably due to the loss of viable
virus during transport, a very slow growth rate of some strains of
adenoviruses, and technical pitfalls in virus isolation because of
toxicity or bacterial contamination.
34 Nevertheless, both
the multiplex and uniplex PCRs failed to detect either viruses in a few
specimens in which virus isolation was reported positive. This could be
due to the occurrence of false-positive virus isolation results due to
cross contamination, although PCR inhibition
35 36 37 cannot
be ruled out. For prevention of the latter, the simple lysis buffer
used as an extraction method, as described
previously,
24 28 38 proved satisfactory because none
of the samples that were culture positive but PCR negative, became
positive when re-extracted by alternative extraction methods (GuSCN and
PEG).
A few reports have claimed the possibility of coinfection by adenovirus
and HSV,
39 40 adenovirus and
C.
trachomatis,
41 or HSV and
C.
trachomatis 42 in cases of KC. None of the samples
tested in this study showed any sign of coinfection by any of the
techniques. However, the possibility of coinfection would favor the use
of the PCR over virus isolation because PCR would detect, for example,
adenovirus DNA in the presence of HSV DNA, whereas cell culture would
lead to production of an early HSV cytopathic effect destroying the
cell cultures and thereby preventing the growth of
adenovirus.
38
A total of 641 (80%) eye swabs remained negative for adenovirus, HSV,
and
C. trachomatis. This apparently high proportion is
partly due to the policy of the local eye hospital. Patients who
present with what, on clinical grounds, is regarded as typical
adenovirus KC are not swabbed. Also, other forms of conjunctivitis such
as those of allergic origin may have contributed to the high percentage
of the negatives in this study. Other possible explanations include the
presence of, for example, adenoviruses that are not amplifiable by the
primer pair ADRJC1/ADRJC2 or the occurrence of
C.
trachomatis strains or other chlamydia that are
plasmid-free.
32 43 44
The ability of even experienced clinicians to diagnose acute
conjunctivitis is limited. The multiplex PCR detected adenovirus DNA in
10 patients who had clinical suspicion of chlamydial KC. The diagnostic
advantages of multiplex PCRs, which include cost-effectiveness and
detection of pathogens that are not suspected clinically, are
highlighted in this study and suggest its usefulness in large and small
centers and in the community to non-specialist staff. Laboratory
charges of the different laboratory techniques used in this study were
approximately $66 for virus isolation, $28 for the Amplicor PCR, $25
for the IDB test, $36 for the multiplex PCR, and $35 for each uniplex
PCR. Thus, for complete viral and chlamydial laboratory investigation
of each specimen, the cost by using standard techniques would be $118
using a combination of virus isolation, Amplicor PCR, and IDB test;
$106 using the uniplex PCRs; and only $36 using the multiplex PCR.
In conclusion, the multiplex PCR described here is sensitive and
cost-effective. The test allows simultaneous screening for the three
pathogens in a single eye swab within a maximum of 7 hours, and the
methodology has the potential to replace routine time-consuming and
costly diagnostic techniques for viral and chlamydial KC.