|
|
||||||||
Articles |
National Center for Infectious Diseases, Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services, Atlanta, GA 30333.
a Address correspondence to this author at: Centers for Disease Control and Prevention, 1600 Clifton Rd., MS G05, Atlanta, GA 30333. Fax 404-639-3123; e-mail TYM2{at}CIDDBD2.EM.CDC.GOV
| Abstract |
|---|
|
|
|---|
Key Words: indexing terms: PCR Streptococcus pneumoniae clinical specimens paraffinized tissues
| Introduction |
|---|
|
|
|---|
There is a lack of rapid, sensitive, and specific tests for diagnosing pneumococcal infections. Historically, identification of S. pneumoniae sepsis was based on obtaining a positive blood culture if other signs and symptoms consistent with the diagnosis were present. The laboratory identification of S. pneumoniae is based on the hemolysis pattern when cultured on blood agar plates and by additional biochemical tests on the cultured organisms: optochin sensitivity, the Quellung reaction, and bile solubility (4). Nonculture methods now available, such as counterimmunoelectrophoresis (CIE) and coagglutination, have failed to show uniform diagnostic sensitivity or specificity even when applied to specimens collected from patients with culture-confirmed bacteremic pneumococcal pneumonia (5)(6)1 The lack of sensitive nonculture-based methods and the limitations of culture often result in only presumptive diagnoses of pneumococcal pneumonia (7). Therefore therapy is frequently empiric and potentially suboptimal (7), and reliably identifying cases of pneumococcal infections for epidemiological studies is difficult.
We applied a new nested PCR for detection of S. pneumoniae from clinical specimens from an outbreak of invasive pneumococcal disease in a child care center (8). In December 1993, two previously healthy infants in a New Mexico child care center developed a severe illness characterized by septic shock and hemorrhage into the skin or internal organs and died. In this outbreak, cultures of blood and spinal fluid from the two patients were negative. Latex agglutination testing and CIE on cerebrospinal fluid (CSF) from the second infant affected suggested infection with S. pneumoniae as the cause of death. Clinical specimens tested included buffy coat, CSF, and autopsy tissue.
This is the first PCR application for detection of pneumococcus in autopsy tissues. Identifying the cause of death in autopsy tissues is important to treat similar cases with appropriate antibiotics, and to determine the etiology of an illness. Several reports describe PCR assays from clinical specimens of blood and (or) buffy coat (9)(10)(11)(12)(13), CSF (14)(15), sputum (16), and ear fluids (17)(18). The assay was critical in identifying the etiology of the deaths in the outbreak.
| Materials and Methods |
|---|
|
|
|---|
detection of s. pneumoniae in blood
A fresh culture of type 19F S. pneumoniae was removed
from a sheep blood agar (SBA) plate (Becton Dickinson Microbiology
Systems) by flooding the plate with 3 mL of sterile isotonic saline and
scraping the colonies off with a sterile inoculating loop. The solution
was vortex-mixed and diluted with sterile saline to a MacFarland
reading of 0.5. This reading is ~1.5 x 108 cells
per mL. Further dilutions in sterile saline were made to adjust the
cell number to 15 cells per µL. Aliquots equivalent to ~1150
organisms were added to 150 µL of fresh blood collected in EDTA tubes
from healthy volunteers. Identical amounts of the cell suspensions that
were added to blood were plated onto fresh SBA plates, incubated
overnight at 37 °C in a CO2 incubator, and the number of
colonies counted the next day for an accurate determination of the cell
number added to blood. Similarly, S. pneumoniae could be
harvested by centrifugation from overnight growth in Pediatric TSB
bottles supplemented with 1 mL of normal human blood and processed in a
like manner as pneumococci grown on SBA plates. There was no difference
in sensitivity of detection of plate-grown vs broth-grown organisms.
The heat detergent method (HD) of Dawson et al. (19) was used to prepare blood samples for PCR. Briefly, the red cells were lysed by hypotonic treatment with ammonium chloride followed by addition of sterile water. The remaining cells were collected by centrifugation, resuspended in 100 µL of detergent buffer (Nonidet P40 and Tween 20), and boiled in a boiling water bath for 30 min. Five to 10 µL were used in a final PCR reaction volume of 50 µL.
dna preparation of streptococci and other bacteria for pcr
After overnight growth on blood agar plates, the cultures of
streptococci or other bacteria were harvested by flooding each plate
with 3 mL of isotonic saline and dislodging the colonies with a sterile
inoculation loop. The suspension was removed to a sterile 1.5-mL
Eppendorf tube and centrifuged at 3000g for 10 min. The
supernatant was removed and the pellet was treated with 50 µL of 0.5
mol/L potassium hydroxide for 30 min at 37 °C. Fifty microliters of
0.5 mol/L Tris, pH 8, was added, and the tubes were boiled in a boiling
water bath for 20 min. The tubes were centrifuged at
16 000g in an IEC MicroMax microcentrifuge (International
Equipment Co., Needham Heights, MA). The supernatants were diluted at
least 1:10 for PCR.
Quantified DNA was prepared by reading the absorbance at 260 nm after extraction on an ABI model 340-A DNA extractor (Perkin-Elmer, Applied Biosystems Division, Foster City, CA). This DNA was used to measure sensitivity of the PCR.
pcr amplification
Samples for PCR were prepared in a class II laminar flow hood, and
amplification and analysis of PCR products were each performed in
separate locations. Reaction volumes of 50 µL containing 10 mmol/L
Tris-HCl, pH 8.8, 75 mmol/L KCl, 2.8 mmol/L MgCl2, 200
µmol/L of each deoxynucleoside triphosphate, 0.1 g/L bovine serum
albumin (Sigma Chemical Co., St. Louis, MO), 1.25 U of Taq polymerase
(Boehringer Mannheim, Indianapolis, IN), 0.2 µmol/L of each outer
primer, and 5 µL of DNA sample were overlaid with one drop of mineral
oil and placed in a Perkin-Elmer Thermal Cycler Model 480
(Perkin-Elmer, Norwalk, CT) for 1 cycle of 94 °C for 4 min, followed
by 30 cycles of 94 °C for 1 min, 62 °C for 30 s, and
72 °C for 1 min. A final 5-min extension at 72 °C concluded the
amplification. The nested or inner PCR reaction mixture was similar to
the first except that it contained 0.1 µL of the product of the outer
PCR and 0.2 µmol/L of each inner primer. The cycling conditions were
identical.
The primers target the pneumococcal autolysin gene sequenced by Garcia et al. (20), which is unique and characteristic of S. pneumoniae (21). A computation at the National Center for Biotechnology Information with the BLAST network service confirmed the uniqueness of the target sequence: The only matches were with the phage HB-3 of S. pneumoniae. There was no human sequence homology.
A750 Outer sense 5'-GGCTACTGGTACGTACATTC-3'
A1145 Outer antisense 5'-AATCAAGCCATCTGGCTCTA-3'
A781 Inner sense 5'-ATCCAAAAGACAAGTTTGAGA-3'
A1055 Inner antisense 5'-CTGGATAAAGGCATTTGATAC-3'
The outer PCR product is 395 bp and the inner product is 274 bp.
Amplification products were separated by electrophoresis through 2.5% agarose gels [1.5% Nusieve GTG agarose (FMC Bioproducts, Rockland, ME) and 1.0% agarose (Bio-Rad, Richmond, CA)] in TrisborateEDTA and were visualized by ethidium bromide fluorescence.
Both the outer and inner PCRs were optimized with the Opti-Prime PCR optimization kit (Stratagene, La Jolla, CA). The PCR product was verified by sequence analysis (National Biosciences, Plymouth, MN).
preparation of clinical specimens
We received CSF, buffy coat, and four paraffinized autopsy
blocks (liver\/spleen\/thymus, diaphragm, and two lung) from one
patient. From the second patient we received three paraffinized blocks:
lung, pancreas\/ heart\/liver, and spleen\/lymph node. Buffy coat
and CSF were prepared for PCR with standard chloroform\/phenol
extraction methods. Buffy coat, CSF, and autopsy tissues from the two
patients in New Mexico were provided by Kurt Nolte of the New Mexico
Office of the Medical Investigator and by the New Mexico Department of
Health laboratory personnel. The procedures followed for human subjects
were in accordance with the ethical standards of the Helsinki
Declaration of 1975, as revised in 1983.
Autopsy tissues were prepared for PCR in a nonrespiratory disease laboratory by using the method of Heller et al. (22). Briefly, 8-µm tissue sections were deparaffinized, and the dried tissue was digested and then extracted with chloroform\/phenol, followed by ethanol precipitation. The pellet was dissolved in sterile, endotoxin-free water, and 5 µL were used in the PCR.
| Results |
|---|
|
|
|---|
|
The detection limit of the nested PCR was 1.25 fg of purified S.
pneumoniae DNA (less than one genome equivalent). An example of
the sensitivity and specificity of the nested reaction is shown in Fig. 1
. The figure illustrates the dilution effect of detecting very
low concentrations of DNA in a sample. Five femtograms and 1.25 fg
amplify, whereas 2.5 fg do not. This is a concentration effect at
limiting dilution of target in the sample. Five femtograms represents
~2 genome equivalents. At a concentration of 1 fg/µL, a random
5-µL aliquot may not contain any genome equivalent, whereas another
5-µL aliquot may contain more than two genome equivalents.
|
Figure 2
shows the detection of S. pneumoniae by nested PCR
in human blood. Four and one-half colony-forming units (CFU) of
S. pneumoniae were detectable. Counted S.
pneumoniae were added to 200 µL of human blood, and the blood
was prepared for PCR with the HD method described in Materials
and Methods. The figure shows the results of varying the volume of
the HD buffer used to boil the final cell pellet on detection by PCR. A
high (too numerous to count) and a low number (1 CFU per µL) of
pneumococci were added to the blood before isolation. There was no
effect on detection by PCR if the final HD buffer volume decreased.
|
Cultures of CSF, blood, and tissues from patients in this study were negative for S. pneumoniae. Gram-positive cocci were detected in blood buffy coat of one patient, and a latex agglutination test of CSF indicated infection with S. pneumoniae. CIE of the CSF specimen was positive for S. pneumoniae (personal communication of D.M. Musher, VA Medical Center, Houston, TX).
Figure 3
shows the results of the second-step PCR amplification of
clinical specimens from the two children in this study. Even though we
optimized the reaction conditions for all primers, the first-step PCR
on the clinical specimens used in this report failed to show bands in
ethidium bromide-stained agarose gels. All of the specimens from both
infants were positive in our second-step PCR; however the CSF-positive
specimen is not shown in this figure. There were four paraffinized
blocks from patient 1, including two from lung, and three blocks from
patient 2. The tissue blocks, sectioned by the Bacterial Diseases
Pathology Laboratory, were deparaffinized and extracted along with the
buffy coat and CSF in a nonrespiratory disease laboratory at CDC. To
confirm our PCR results, the tissue blocks were given to the Viral
Diseases Pathology Laboratory where new sections were made,
deparaffinized, and digested. We found identical PCR positive results
on all of the tissue specimens sectioned and prepared from two
different laboratories (data not shown).
|
| Discussion |
|---|
|
|
|---|
We designed a nested PCR to detect pneumococci in respiratory infection. The target sequence was from the autolysin gene common to all S. pneumoniae. A nested strategy was used to increase specificity and sensitivity of detection of target in clinical specimens. The nested reaction is a two-step PCR. One percent or less of the first-step product is amplified in the second-step PCR, and the second-step primers target a sequence internal to the first-step sequence. Clinical specimens often contain inhibitors to PCR even after purification steps such as chloroform\/phenol extraction. The two-step PCR often circumvents the effects of inhibitors in clinical specimens because while the first-step product may be too small for detection by ethidium bromide staining, enough product is synthesized for amplification and detection in the second step or nested reaction.
The detection limit of the nested PCR was equivalent to the amount of target contained in one bacterium as determined by using purified DNA or 4.5 CFU in blood. All of the type strains of S. pneumoniae were amplifiable with the expected size products. The most common forms of streptococci found in respiratory airways that could cross-react are S. viridans and group C; however, PCR products were not detected with any of these organisms tested, nor with other respiratory pathogens that might be present in clinical specimens of interest.
The PCR we developed was used to identify the etiology (S. pneumoniae) of an unknown illness causing the deaths of two children in an Albuquerque child care center. One of the two infant deaths occurred 6 weeks before the investigation, and autopsy tissues were the only specimens remaining for one of the infants. This method was crucial to linking the two deaths to one etiology.
Because this study involved only two cases, the general use of PCR to retrospectively diagnose pneumococcal disease from paraffinized tissues must be interpreted with caution. Asymptomatic carriage of S. pneumoniae in up to 20% of adults requires the exclusion of detection of pneumococcus in control tissues. A prospective study of detection in control tissues removed and handled sterilely at autopsy is required to establish the efficacy of this test as a general method.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
D. Llull, R. Lopez, and E. Garcia Characteristic Signatures of the lytA Gene Provide a Basis for Rapid and Reliable Diagnosis of Streptococcus pneumoniae Infections J. Clin. Microbiol., April 1, 2006; 44(4): 1250 - 1256. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Sa-Leao, A. S. Simoes, S. Nunes, N. G. Sousa, N. Frazao, and H. de Lencastre Identification, prevalence and population structure of non-typable Streptococcus pneumoniae in carriage samples isolated from preschoolers attending day-care centres Microbiology, February 1, 2006; 152(2): 367 - 376. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Moreno, E. Hernandez, O. Sanabria, and E. Castaneda Detection and Serotyping of Streptococcus pneumoniae from Nasopharyngeal Samples by PCR-Based Multiplex Assay J. Clin. Microbiol., December 1, 2005; 43(12): 6152 - 6154. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |