Quantification of Chlamydia trachomatis Elementary Bodies in Urine by Ligase Chain Reaction

Clinical specimens.First-catch urine was obtained in an ongoing study of commercial sex workers in Madagascar (K. Van Damme, L. Raharimalala, N. Ratsimbazafy, et al., Abstr. Int. Congr. Sex. Transm. Infect., p. 180, 2001). Informed consent was obtained from all patients with approval of the Committee for the Protection of Human Subjects at the University of North Carolina at Chapel Hill and the Ethical Review Board of the National HIV Reference Laboratory in Madagascar.

Preparation of C. trachomatis stock. C. trachomatis was cultivated as previously described with slight modification (3). Briefly, McCoy cells (ATCC 1696) were grown to near confluence in Dulbecco's modified medium supplemented with 5% fetal bovine serum, 25 mM HEPES, 2 mM l-glutamine, and gentamicin (10 μg/ml) at 37°C in 5% CO₂. McCoy cell division was arrested with cycloheximide as described previously (3). C. trachomatis serovar E elementary bodies (EB) (generously provided by Priscilla B. Wyrick, East Tennessee State University) were added and incubated for 48 to 72 h at 35°C. The contents of several culture flasks were pooled, sonicated for 5 min, and centrifuged at 500 × g for 10 min at 4°C to remove cellular debris. The supernatant was centrifuged at 10,000 × g for 30 min at 4°C; the resulting pellet was resuspended in buffer containing sucrose, phosphate, and glutamate (2-SPG). EB were isolated from this crude C. trachomatis preparation by purification over a Renografin gradient. Purified EB were suspended in 2-SPG, aliquoted, and stored at −80°C.

Quantification of C. trachomatis stock solution.Purified EB were thawed at 37°C and sonicated on ice for 5 min. The concentration of C. trachomatis EB was determined by direct microscopic count using a fluorescein-labeled monoclonal antibody as previously described (8). Briefly, 10 μl of a 1:4 dilution of EB stock solution was placed into 6-mm-diameter wells on a Teflon-coated slide (catalog no. 63424-06; Electron Microscopy Sciences), air dried, and fixed with 70% ethanol-30% acetone for 10 min. Wells were incubated with 15 μl of fluorescein isothiocyanate-labeled antichlamydia antibody (chlamydia confirmation kit [catalog no. 8H0194]; Wompole Labs) for 30 min at 37°C; slides were rinsed and allowed to air dry in the dark.

Immediately after staining, slides were viewed using a Zeiss LSM210 confocal laser scanning microscope in epifluorescence mode with a magnification of ×1,000. Nine fields were counted for each well; seven individual wells on four independent slides were examined. The area per high-power field (hpf) was determined using a stage micrometer. The mean number of organisms per hpf was determined for each well and averaged for seven wells. The stock concentration of EB per milliliter was determined using the following formula: (average no. of EB per hpf) × (area of individual well) × (reciprocal dilution of stock) × 1,000/(area per hpf) × (volume of sample [in microliters]).

Preparation of quantitative standards.For quantitative LCR (qLCR), three overlapping series of twofold dilutions of C. trachomatis stock solution were prepared in pooled urine from uninfected volunteers. Several diluents were tested by using the standard LCR urine sample preparation procedure (Abbott LCx probe system); pooled urine resulted in higher rates by LCR than did 2-SPG or phosphate-buffered saline. Low-range standards contained 3.2 × 10¹ to 1.0 × 10³ EB/ml, medium-range standards contained 5.1 × 10² to 3.3 × 10⁴ EB/ml, and high-range standards contained 1.6 × 10⁴ to 1.0 × 10⁷ EB/ml. When necessary, a medium-low-range standard with 1.3 × 10² to 4.1 × 10³ EB/ml or a high-medium-range standard containing 4.1 × 10³ to 1.0 × 10³ EB/ml was used. C. trachomatis stock solutions and LCR preparations that were frozen and thawed up to 20 times prior to qLCR testing showed no significant decrease in LCR rate. To maximize uniformity of individual test runs, an initial stock was quantified microscopically, aliquoted into multiple-use vials, and frozen at −80°C. Dilutions were made from a freshly thawed stock solution and prepared according to the LCR urine protocol. For each qLCR run, an appropriate standard dilution series was amplified with clinical specimens.

qLCR.Clinical urine specimens were screened using the Abbott LCx probe system according to the manufacturer's instructions. For qLCR, positive urine specimens were prepared according to the standard LCR protocol and amplified alongside a medium-standard series for 30 cycles of the otherwise-standard thermocycler profile. The log₂ of the LCR rate was plotted against the concentration of each standard, and linear regression was used to generate the equation of the line. A valid qLCR run was defined as one for which there was an increasing LCR sample rate for each standard in the series and the R² of the regression line was ≥0.900. Chlamydia concentrations in clinical specimens were calculated from the equation of the standard line.

To verify the reproducibility of LCR with altered cycle numbers, five preparations were made from a clinical urine sample that tested positive for C. trachomatis by the standard LCR. Each preparation was amplified using 30 cycles, and one preparation was tested five different times. There was minimal variation between preparations and between tests with the same preparation.

Samples that fell outside the linear range of the medium-standard series were retested using the low-standard series at 35 amplification cycles or the high-standard series at 25 cycles, as appropriate. In rare instances, LCR sample rates were not squarely within the medium-standard or the subsequent series. These samples were analyzed with an intermediate series, 33 cycles for the medium-low-range standard series or 28 cycles for the high-medium-range standard series.

Statistical analysis.Descriptive statistics and linear and polynomial regression were performed using SigmaStat software (version 2.03; SPSS, Inc.).

qLCR of clinical specimens.Urine was obtained in an ongoing study of commercial sex workers in Madagascar as described in Materials and Methods. Specimens from 1,000 women were screened for C. trachomatis using standard LCR. One hundred fifty-eight specimens were confirmed positive for C. trachomatis by standard LCR and underwent qLCR testing. In order to demonstrate the need to modify the qualitative protocol for quantitative analysis, the relationship between the output from the standard LCR assay and the qLCR results, the ratio of the sample rate to the cutoff value (S/CO, an internally standardized value calculated automatically by the Abbott LCx analyzer), was plotted against C. trachomatis concentration for each specimen. Incremental polynomial regression analysis suggested that the relationship followed second-order kinetics, with an R of 0.615. The scatter plot and regression line for the quadratic equation are shown in Fig. 2. The relationship between S/CO and organism burden was approximately linear at lower C. trachomatis concentrations (less than 1,000 EB/ml of urine) and plateaued at higher concentrations. The highest concentration for samples with an S/CO value of ≤2 was 140 EB/ml of urine. This demonstrates that the degree of amplification provided by the qualitative test shifted the values for the higher concentrations to the asymptotic portion of the sigmoid curve. The higher correlation coefficient of the qLCR method suggests that the quantitative standards provided a better estimate for quantification than the internal standards of the qualitative LCR assay.

• Open in new tab
• Download powerpoint

FIG. 2.

Comparison of standard LCR S/CO values and qLCR results from clinical urine specimens from Malagasy women. The solid line is the second-order-regression line (R = 0.615); dotted lines represent the 95% confidence intervals of the regression line.

The distribution of C. trachomatis concentrations in urine from this population is shown in Fig. 3. Fifteen samples were confirmed positive for C. trachomatis by standard LCR but were below the quantifiable limit of qLCR. The median concentration of quantifiable samples was 2.17 × 10³ EB/ml (range = <3.2 × 10¹ to 2.19 × 10⁵ EB/ml). Quantifiable samples displayed an apparent bimodal distribution. Thirty-seven percent (53 of 143) of samples were in the lower peak (range = 3.2 × 10¹ to 1.02 × 10³ EB/ml; median = 3.2 × 10² EB/ml), whereas 63% (90 of 143) had relatively higher concentrations (range = 1.1 × 10³ to 2.19 × 10⁵ EB/ml; median = 7.39 × 10³ EB/ml).

• Open in new tab
• Download powerpoint

FIG. 3.

Histogram of the distribution of qLCR values obtained with clinical urine specimens from Malagasy women. The numbers above the bars indicate the number of specimens with values in the range of concentrations indicated on the x axis.