B. Lam¹ and Y. Haj-Ahmad^1,2. ¹Norgen Biotek Corp., Thorold, ON, CANADA, ²Brock University, St. Catharines, ON, CANADA

 

 

Background: 

 

Urinary tract infection is a common disease that is caused by bacteria including E coli, Staphlococcus aureus, and Chlamydia trachomatis. The traditional method of urine culture identification is relatively time-consuming. On the other hand, DNA/PCR-based approaches provide results in a shorter amount of time. Also, RNA and protein markers have been used for bacterial identifications. However, these molecular approaches have a few disadvantages: 1) The methods are sensitive to sample contamination due to handling. 2) Inhibitors are usually co-purified with the molecules and affect applications like PCR. 3) DNA, RNA and proteins are usually isolated separately and it becomes problematic when the sample amount is limited. The goals of our research were to develop a method that isolates high quality DNA, RNA and protein simultaneously from the same sample such that sample handling is minimized and hence sample contamination is reduced.  

 

Methods:

 

Urine samples from healthy individuals were spiked with various amounts of Gram positive (Bacillus cereus) or Gram negative (E coli) bacteria. The urine bacteria were preferentially collected, and lysed. DNA, RNA and proteins from the same lysate were sequentially purified using silicon carbide loaded columns. The purified DNA, RNA and proteins were subjected to qPCR, qRT-PCR and SDS-PAGE to detect markers for each bacterium. The purified DNA was subjected to PCR-based sequencing to confirm the identity of the bacteria. 

 

Results:

 

Using primers against the E coli 16s rRNA hypervariable region, it was demonstrated that E coli DNA and RNA could be isolated from less than 1000 cfu (<15000 cfu per mL in a healthy male) in 1 mL of urine by qPCR and qRT-PCR, respectively. Sequencing using the PCR product generated from DNA isolated using a general 16s rRNA primer confirmed the identity of each spiked bacterium. Proteins resolved on SDS-PAGE showed high integrity without degradation. 

 

Conclusion: 

 

A new method based on silicon carbide loaded mini-columns was shown to isolate concurrently high quality bacterial DNA, RNA and

protein from urine applicable to downstream procedures for species detection.

 

 

• Urinary tract infection (UTI) is a common disease that is caused by bacterial infection of any part of the urinary tract. Major categories of UTI are determined by where the infection occurs and include: bacteriuria (urine), cystitis (bladder), prostatitis (prostate gland), pyelonephritis (kidney) and urethritis (urethra). Globally, it is estimated that over 150 millions cases of UTIs occur every year (1).

 

• The most common species causing UTIs include Escherichia coli, Stapholococcus aureus, and Chlamydia trachomatis. Traditionally, the identification of the causal agent of an UTI involves culturing the urine, however this method is very time-consuming. In addition, an extensive amount of work may be required to identify a specific bacterial strain when using this method. In light of such drawbacks, molecular approaches have now become increasingly popular for bacterial identification in UTI diagnosis. 

 

• All three major biomolecules - DNA, RNA and proteins - have been successfully used in UTI detections. For example, the 16S rRNA gene and the 16S-23S rRNA spacer region of the bacterial genomic DNA have been commonly targeted for PCR amplifications of bacterial DNA (2, 3). Similarly, the bacterial major outer membrane protein (MOMP) is usually the target protein in many commercially available antibody-based protein detection products (4). Recently, there has been an increase in the use of RNA markers in addition to DNA. This is particularly important during UTI treatment monitoring, as the isolated genomic DNA may have originated from non-viable bacteria but RNA is usually isolated and detected exclusively from actively-dividing bacteria (5).

 

• While molecular approaches have contributed significantly to the diagnosis of urinary bacterial infections, a few drawbacks still exist. The aforementioned methods are sensitive to sample contamination, particularly during sampling handling. DNA, RNA or proteins are  usually isolated separately and this becomes problematic when the urine sample amount is limited. Moreover, inhibitors are commonly co-purified and may affect some detection methods such as PCR and RT-PCR.

 

• The goals of our research were thus to develop a sample preparation method for UTI detection by:

 

1. Isolating bacterial DNA, RNA and proteins sequentially from the same urine sample, without splitting the sample and with a minimal amount of handling.

2. Isolating bacterial DNA, RNA and proteins from a limited amount of urine.

3. Isolating high quality bacterial DNA, RNA and proteins that are not inhibited for downstream detection methods.

 

 

• Input was 1 mL urine samples spiked with 1.5 x 10⁶, 1.5 x 10⁵, 1.5 x 10⁴, 1.5 x 10³ and 150 cfu per mL of E. coli or 1.5 x 10⁶   cfu per mL of B. cereus

 

• Bacterial cells in the urine were pelleted and lysed. Lysate was loaded onto silicon carbide spin columns and the flowthrough was collected and set aside for protein purification. The column was washed and the bound RNA was eluted, and the column was washed a second time and the bound genomic DNA was eluted. The column was then regenerated and activated, and the flowthrough containing the proteins was loaded onto the column. The bound proteins were then washed and eluted. This procedure is outlined in Figure 1. For more details about the procedure please refer to the manual provided with Norgen’s RNA/DNA/Protein Purification Kit (Cat # 23500).

 

• RNA samples (7.5 μL out of each 50 μL elution) were run on 1.5% formaldehyde agarose gels, DNA samples (12.5 μL out of each 100 μL elution) were run on 1.2% agarose gels and protein samples (20 μL out of each 100 μL elution) were run on 10% SDS-PAGE gels for visual analysis.  

 

• PCR and qPCR (for gDNA) and RT-PCR and qRT-PCR (for RNA) reactions were performed using the purified molecules from the 1.5 x 10⁶ spiking and primers specific for E. coli, B. cereus, or with general detection primers using a Bio-Rad iCycler Thermal Cycler.  7.5 μL out of each 20 μL PCR reaction was run on an agarose gel for visual detection of the species specific amplification.

 

• PCR and qPCR (for gDNA) and RT-PCR and qRT-PCR (for RNA) reactions were performed using the purified molecules from the E. coli spiking and primers specific for the 16S rRNA of E. coli using a Bio-Rad iCycler Thermal Cycler. 7.5 μL out of each 20 μL PCR  reaction was run on an agarose gel for visual detection of the sensitivity, and Ct charts and graphs were also generated to show the sensitivity and linearity of the detection.

 

 

 

 

Figure 2: Isolation of High Quality Bacterial Genomic DNA, RNA and Proteins from 1 mL of Urine. 1 mL urine samples were spiked with 1.5 x 10⁶ E. coli or B. cereus cells and then subjected to 3-in-1 isolation of RNA, DNA and proteins using silicon carbide columns and the procedure outlined in the Methods section. Panel A is a 1.2% formaldehyde agarose gel showing the isolated RNA, Panel B is a 1.2% agarose gel showing the isolated genomic DNA, and Panel C is a 10% SDS-PAGE gel showing the isolated proteins. In each case, Lanes 1 and 2 correspond to urine samples spiked with E. coli., Lanes 3 and 4 correspond to urine spiked with B. cereus, and Lanes 5 and 6 are plain urine. In all cases it can be seen that the isolated molecules are of the highest quality, with no signs of any degradation. 

 

 

Figure 3: Purified Urine Bacteria DNA and RNA are Suitable for Various Identifiation Methods Including PCR and RT-PCR. 1 mL urine samples were spiked with 1.5 x 10⁶ E. coli or B. cereus cells, and the genomic DNA and RNA were isolated from the urine samples are described in the Methods section. The isolated genomic DNA was then subjected to species-speific PCR amplification, and the results can be seen in the three different 1% agarose gels in Panel A. Gel A coresponds to the results when E. coli-specific primers were used, Gel B shows the PCR products when B. cereus primers were used and Gel C shows the results when general bacteria detection primers were used. As it can be seen, the E. coli specific primers detected the DNA isolated from both E. coli alone (Gel A, Lanes 1 and 2) and urine spiked with 1.5 x 10⁶ E. coli (Gel A, Lanes 3 and 4), while the B. cereus-specific primers detected DNA isolated from urine spiked with 1.5 x 10⁶ B cereus (Gel B, Lanes 5 and 6). The general primers detected all the bacterial DNA present (Gel C, Lanes 1-6). The control of plain urine showed no amplification (All gels, Lanes 7 and 8). The M lanes contain DNA markers. Panel B shows the results when the isolated bacterial RNA was used as a template in RT-PCR reactions using the same species-specific primers. The E. coli specific pirmers detected the RNA isolated from urine spiked with E. coli (Gel D, Lane 1), and the B. cereus specific primer amplified the RNA isolated from urine spiked with B. cereus (Gel E, Lane 3) in cases where the reverse transcriptase was added (+ lanes). The - lanes indicate samples in which reverse transcriptase was not added, and these lanes acted as controls to show that no genomic DNA contamination was present in the RNA samples. Thus the isolated urine bacterial DNA and RNA can be used in PCR and RT-PCR reactions to identify the species of bacteria that are present, and no inhibitors to the downstream applications are present

 

 

Figure 4: High Sensitivity of Bacterial Genomic DNA Isolation from 1 mL Urine Samples. Increasing amounts of E.coli were spiked into 1 mL urine samples and the bacterial DNA was isolated as outlined in the Materials section. The DNA was then subjected to PCR using E.coli-specific primers which recognize the 16S rRNA gene. The resulting PCR products were run on a 1.5% agarose gel for visualization, and can be seen in Panel A above. The number of E. coli cells used to spike the urine is 1.5 times the number indicated above the gel. It can be seen that the PCR reaction resulted in the expected 544 bp PCR product in all cases, even when as little as 150 E. coli cells were used. This is an important observation, as a healthy individual generally has < 15,000 bacterial cells/mL of urine, and thus this method is sufficiently sensitive to isolate and detect bacterial DNA from even this low amount of bacteria.   Small aliquots of the urine bacterial DNA were also amplified in a real-time PCR reaction, and the Ct graph generated can be seen in Panel B. The number of E. coli cells used to spike the urine is again 1.5 times the number indicated on the graph. Again, it can be seen that DNA can be isolated and detected from as little as 150 E. coli cells in 1 mL of urine. Thus this 3-in-1 method is extremely sensitive for the isolation of urine bacterial DNA. 

 

 

Figure 5: High Sensitivity of Bacterial RNA Isolation from 1 mL Urine Samples. 1 mL urine samples were spiked with either 15,000, 1,500 or 150 E coli cells. Bacterial RNA was isolated from these urine samples, as well as from unspiked urine samples, using the procedure outlined in the Materials section. The RNA was then used as the template in a RT-PCR reaction, and the results can be seen in Panel A. It can be seen that in all cases the isolated bacterial RNA could be detected in the RT-PCR reaction, as evidenced by the presence of the 544 bp band in all the lanes where the reverse transcriptase was added (+ lanes). The - lanes indicate samples in which reverse transcriptase was not added, and these lanes acted as controls to show that no genomic DNA contamination was present in the RNA samples. Real time RT-PCR was also performed on the RNA samples, and the corresponding Ct graph can be seen in Panel B. Again, the RNA could be isolated and detected from all the samples, indicating the high sensitivity of this procedure for isolating bacterial RNA from urine (red lines). RNA could even be isolated and detected from 1 mL of regular non-spiked urine (blue line). 

 

1. A procedure based on the use of silicon carbide columns was developed that allows for the sequential isolation of bacterial RNA, genomic DNA and proteins from the same urine sample without any splitting of the lysate (Figure 1).

 

2. The bacterial RNA, DNA and proteins isolated are of the highest quality, and show no signs of degradation (Figure 2).

 

3. The isolated RNA and DNA contain no inhibitors, and can be used directly in downstream applications including PCR, RT-PCR and real-time PCR (Figures 3, 4, 5).

 

4. The isolated RNA and DNA can be used successfully in PCR and RT-PCR reactions to detect and identify different bacterial species found in1 mL urine samples (Figure 3).

 

5. There is no genomic DNA contamination in the purified RNA fraction, as evidenced by a lack of PCR product when RT-PCR is performed without the addition of reverse transcriptase (Figures 4 and 5, - Lanes)

 

6. The developed method is sensitive enough to isolate and detect RNA and gDNA from as little as 150 E. coli cells present in 1 mL of urine. Normal humans generally have < 15,000 bacterial cells/mL of urine, thus this procedure is sensitive enough to isolate the RNA, DNA and proteins from even this small amount of bacteria (Figures 4 and 5).

 

 

1. Stamm, W.E. and Norrby S.R. 2001. Urinary tract infections: disease panorama and challenges. J. Infect. Dis. 183(Suppl. 1): S1-S4.

 

2. Madico G., Quinn T.C., Boman J. and Gaydos C.A. 2000. Touchdown enzyme time release-PCR for detection and identification of Chlamydia trachomatis, C. Pneumoniae, and C. Psittaci using the 16S and 16S-23S spacer rRNA genes. J. Clin. Microbiol. 38: 1085-1093.

 

3. Jurstrand M., Jensen J.S., Fredlund H., Falk L. and Mölling, P. 2005. Detection of Mycoplasma genitalium in urogenital specimens by real-time PCR and by conventional PCR assay. J. Med. Microbiol. 54: 23-29.

 

4. Boman J., Gaydos C., Juto P., Wadell G. and Quinn, T.C. 1997. Failure to detect Chlamydia trachomatis in cell culture by using a monoclonal antibody directed against the major outer membrane protein. J. Clin. Microbiol. 35: 2679-2680.

 

5. Takahashi S., Shimizu T. Takeyama K., Furuya R., Koroku M., Tanda H., Nishimura M., Iwasawa A., Yoshio H., Iton N. and Tsukamoto T. 2003. Efficacy of an RNA detection test kit in the diagnosis of genital chlamydial infection. J. Infect. Chemother. 9: 90-92.

 

 

The authors would like to thank NSERC for their financial support (IRF) of this project. The authors would also like to thank Pam Roberts for her help in preparing this poster.