High Throughput Multiplex STR Systems Without Instrumentation

Ann M. Lins, Cynthia J. Sprecher, Katherine A. Micka, Dawn Rabbach, Jeffrey Bacher and James W. Schumm
Promega Corporation, 2800 Woods Hollow Road, Madison, WI 53711 USA

× Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø × Ø

INTRODUCTION

Short tandem repeat (STR) loci (1-4) are becoming widely used in numerous applications, including forensics and paternity, for many reasons. First, amplification of STRs require minimal amounts of DNA and many rapid DNA purification methods are compatible with amplification. STR loci have discrete and easily assigned alleles. The small, defined size ranges allow for the development of multiplexes. Allelic ladders can be used for quick, accurate sizing of alleles. The detection format is non-radioactive. The laboratory hands-on time from sample to result is short, allowing fast turn around times. And lastly, there is the potential for automation, which is already occurring with robotics, fluorescent instrumentation, and computerized allele scoring.

We describe the development of a rapid, efficient, nonisotopic, and inexpensive method for DNA analyses based upon the amplification of polymorphic STR loci. We have achieved high throughput analysis with STR loci for forensic analysis and paternity determination using manual separation and silver stain detection to avoid the need for specialized or expensive equipment. Our approach includes the development of thirteen STR systems in combination with allelic ladders (i.e., size standards containing many or all of the existing alleles) for rapid typing of each locus.

Six of the STR loci have been incorporated into two multiplexes each containing three loci. The first triplex includes the loci CSF1PO, TPOX, TH01 (CTT), while the second contains F13A01, FESFPS, vWA (FFv). This approach minimizes the amount of material required while increasing the efficiency of analysis. An additional triplex containing the loci D16S539, D7S820, D13S317 is currently in development and is included in this discussion. Monoplex systems have been developed for each of the loci present in one of the triplex sets, plus for the loci D5S818, F13B, LPL and HPRTB.

We have collaborated with other laboratories to determine the frequency of alleles for nine STR loci in three or more population groups each containing at least 200 individuals. In addition, the CTT multiplex was validated for use in a study involving five laboratories. This system was tested for amplification quality with 0.5- 250 ng of template, for sensitivity to variation of the annealing temperature, and for the ability to detect and interpret DNA mixtures.

All materials necessary for amplification except Taq DNA Polymerase were supplied in the GenePrint™ STR Multiplex System-CSF1PO, TPOX, TH01 and the GenePrint™ STR Multiplex System-F13A01, FESFPS, vWA (Promega, Madison, WI). SilverSTR™ III multiplex (D16S539, D7S820, D13S317, developed in this work). All amplification and detection protocols are provided in the GenePrint™ STR Systems Technical Manual (Promega, Madison, WI). Amplified fragments are separated in a 4% denaturing polyacrylamide gel with 7M urea and 0.5X TBE as the running buffer. The gel is temporarily bound to one glass plate to simplify handling during the silver stain procedure. The denaturing polyacrylamide gels are subjected to electrophoresis for 60 to 75 minutes. The silver stain procedure of Bassam et al. (5) is then employed. A permanent record and multiple copies may be obtained using Automatic Processor Compatible Film (Promega, Madison, WI). The entire process generally takes less than one day and does not require expensive instrumentation.

In January of 1996, TWGDAM conducted a survey of its members to determine the most desirable STR locus characteristics and criteria to be used in selecting STR loci. The members that responded indicated that multiplex STR systems are important for high throughput applications. In addition, each locus should have moderate polymorphism with at least 6-8 alleles and the STR loci should be easy to amplify and designate. In a multiplex system, the loci should not overlap. The alleles within a locus and between loci should be balanced (i.e., similar peak heights or intensity). Lastly, allelic ladders are very useful tools that should be included in the analysis. These results were very reassuring as our development of STR loci and multiplex systems have included all of these features.

In the selection and development of STR loci, we have used the following criteria: heterozygosity, amplification yield and sensitivity, and simplicity of interpretation. In our evaluation of STRs, loci with numerous 1 and 2 base variants that are difficult to interpret have been excluded. In addition, loci which have many PCR-generated artifacts such as repeat slippage (6, 7) have also been rejected. For multiplex development, we consider the chromosome location and allele size range such that none of the loci overlap. In addition, the loci must have compatible amplification protocols. When the loci are combined into one reaction, there must not be any cross-locus amplification artifacts. The development of quality loci and multiplex systems takes considerable time and effort.

Table 1 lists the fourteen loci we have developed for silver stain detection, as well as fluorescence detection (Schumm et al., this volume)(8). Amelogenin is not an STR, but may be used as a sex identification locus which displays a 212 base X-specific band and a 218 base Y-specific band. The other thirteen STR loci have moderate heterozygosities ranging from 63 to 83 percent in the tested racial groups. All the loci are located on different chromosomes except for CSF1PO and D5S818 which are both on the long arm of chromosome 5. The size range for each locus is well defined to allow for multiplex development. All of the loci contain true tetranucleotide repeats, except for Amelogenin. Within this group, two common microvariants have been identified and characterized. Locus TH01 contains an allele 9.3 microvariant which we published (9) as missing a single adenine in the seventh copy of the AATG repeat in what would otherwise be a normal allele 10. Another common microvariant is allele 3.2 in the locus F13A01. In 1994 we published that the F13A01 allele 3.2 is actually missing two bases flanking the repeat region (10). Additional rare sequence modifications have been observed in the loci TH01, CSF1PO, and HPRTB.

Our laboratory has completed the development of two STR multiplex systems for silver stain detection - the CSF1PO, TPOX, TH01 (CTT) Multiplex and the F13A01, FESFPS, vWA (FFv) Multiplex. A third multiplex containing the loci D16S539, D7S820, D13S317 is currently in development. The work we have completed has been published in two BioTechniques articles. Sprecher et al. (11) discusses the selection and analysis of STR loci and Lins et al. (12) focuses on multiplex development for silver stain and fluorescence detection.

Figure 1 displays the CTT multiplex, the first triplex developed for silver stain detection. Samples in the numbered lanes have been amplified simultaneously at the loci CSF1PO, TPOX, and TH01. Each allele differs from the next by four bases due to the addition or deletion of a single copy of the tetranucleotide repeat unit. Lanes labeled L contain a mixture of allelic ladders for all three loci. The allelic ladders allow for the rapid and accurate calling of alleles.

In 1994, Promega collaborated with the FBI Academy, Palm Beach Sheriff’s Crime Lab, The Blood Center of Southeastern Wisconsin, and Berkeley DNA Laboratory to validate the CTT multiplex (13). As part of the validation, the sensitivity of the multiplex was determined. Figure 2 displays the amplification of as little as 0.5ng template DNA. Amplification of the CTT multiplex is very robust as displayed in Figure 3. The same three DNA samples were amplified using the CTT multiplex with different annealing temperatures. The reactions in the center were completed using the recommended annealing temperature of 64°C. The samples on the left were amplified using a 62°C annealing temperature, while the samples on the right were amplified using an annealing temperature of 66°C. Little to no difference is seen in product yield and the alleles are identical. In addition, each of the five contributing laboratories determined the triplex loci alleles for twenty DNA samples which had been prepared using four different DNA purification procedures. Results with monoplex systems were compared with the multiplex system. Use of the Perkin-Elmer model 480 and model 9600 thermal cyclers were also compared in the study. The validation work for the CTT multiplex was published in the Journal of Forensic Sciences (13).

The second multiplex developed for silver stain detection is the FFv system. Figure 4 displays twenty samples, each amplified simultaneously at the loci F13A01, FESFPS, and vWA. Once again, allelic ladders were prepared and included to simplify analysis. Figure 5 shows the sensitivity of the FFv multiplex. The numbered lanes contain 25ng, 10ng, 5ng, 2ng, 1ng, 0.5ng, and a negative control. In Figure 6, the FFv multiplex was used to amplify the same three samples but with different annealing temperatures. The reactions in the center were done using the recommended annealing temperature of 60°C. The samples on the left were amplified using a 58°C annealing temperature, while the samples on the right were done at an annealing temperature of 62°C.

We are continuing to evaluate additional loci and multiplex combinations for silver stain detection. Figure 7 shows the SilverSTR™ III multiplex which is still in development. Twenty-one DNA samples were amplified at the loci D16S539, D7S820, and D13S317. In observing this multiplex, it has very few PCR-generated artifacts. Figure 8 displays the sensitivity of SilverSTR™ III. The amplification of all three loci is successful when using as little as 0.5ng template DNA. Figure 9 displays the evaluation of annealing temperatures for the SilverSTR™ III multiplex. The same three samples were amplified using three different annealing temperatures. The reactions in the center were completed using the recommended annealing temperature of 60°C. The samples on the left were amplified using a 58°C annealing temperature, while the samples on the right were amplified using an annealing temperature of 62°C.

To generate population data, we have collaborated with other laboratories to determine the frequency of alleles for all nine loci contained within the three multiplexes described in this paper. The collection of data included the analysis of more than two hundred individuals from each of the three major racial groups in the United States.

Table 2 shows the matching probabilities (14) for each multiplex system in various populations. By employing all three multiplexes, the nine loci provide a matching probability of 1 in 4.93 billion, 1 in 1.05 billion, and 1 in 1.83 billion for the three racial groups, respectively.

For paternity analysis, the typical paternity index (PI) (15) using a particular system (e.g., STR multiplex) is calculated to determine the genetic odds in favor of paternity. A PI of 100 indicates a 99% probability of paternity. Table 3 shows the typical paternity index for each multiplex system in various populations. When using all three multiplexes for a total of nine loci, the typical paternity index is 985 for African-Americans, 721 for Caucasian-Americans, and 560 for Hispanic Americans. Another calculation used in paternity analysis is power of exclusion (15). Table 4 lists the power of exclusion for each multiplex system in various populations. The power of exclusion when using all three multiplexes is above 99.8% for all three racial groups.

Our laboratory has carefully selected and developed STR loci which have minimal genetic artifacts (e.g., microvariants) and minimal PCR-generated artifacts (e.g., repeat slippage). The STR multiplex systems containing non-overlapping loci have been developed to display excellent amplification yield and quality. The described multiplex systems allow for the high throughput analysis of STR loci using manual separation and silver stain detection to avoid the need for specialized or expensive equipment. Population data and additional validation work has been generated to demonstrate the usefulness of these STR multiplex systems for forensic analysis and paternity determination.

We would like to thank Steve D. Creacy and Robert A. Bever for their collaboration on development of the population data described in this work.

REFERENCES

Figure 1. Amplification of genomic DNA samples using the CSF1PO, TPOX, TH01 (CTT) Multiplex. The samples in the numbered lanes have each been amplified simultaneously at the loci CSF1PO, TPOX, and TH01. The amplified fragments were separated in a 4% denaturing polyacrylamide gel and detected by silver staining.

Lanes labeled L contain a mixture of allelic ladders for all three loci. Each allelic ladder is labeled to the right with the number of copies of the repeated sequence contained within its corresponding largest and smallest alleles.

Figure 2. Amplification sensitivity using the CSF1PO, TPOX, TH01 (CTT) Multiplex. K562 DNA ranging from 25ng to 0.5ng was amplified using the CTT Multiplex, separated in a 4% denaturing polyacrylamide gel, and detected by silver staining. Lanes 1-6 contain amplified K562 DNA using 25, 10, 5, 2, 1, and 0.5ng template DNA, respectively; lane 7 contains a negative control amplification reaction (i.e., no template DNA added). Lanes labeled L contain a mixture of allelic ladders for the multiplex.

Figure 3. Effects of varying the annealing temperature in the amplification protocol. The CSF1PO, TPOX, TH01 Multiplex was used to amplify three DNA samples using the recommended 64°C annealing temperature or with an annealing temperature two degrees above or two degrees below the recommended temperature. Lanes labeled 1, 2, 3 contain the amplified DNA samples; lanes labeled N contain the negative control amplification (i.e., no template DNA added); lanes labeled L contain a mixture of the allelic ladders for the multiplex.

Figure 4. Amplification of genomic DNA samples using the F13A01, FESFPS, vWA (FFv) Multiplex. The samples in the numbered lanes have each been amplified simultaneously at the loci F13A01, FESFPS, and vWA. The amplified fragments were separated in a 4% denaturing polyacrylamide gel and detected by silver staining.

Lanes labeled L contain a mixture of allelic ladders for all three loci. Each allelic ladder is labeled to the right with the number of copies of the repeated sequence contained within its corresponding largest and smallest alleles.

Figure 5. Amplification sensitivity using the F13A01, FESFPS, vWA (FFv) Multiplex. K562 DNA ranging from 25ng to 0.5ng was amplified using the FFv Multiplex, separated in a 4% denaturing polyacrylamide gel, and detected by silver staining. Lanes 1-6 contain amplified K562 DNA using 25, 10, 5, 2, 1, and 0.5ng template DNA, respectively; lane 7 contains a negative control amplification reaction (i.e., no template DNA added). Lanes labeled L contain a mixture of allelic ladders for the multiplex.

Figure 6. Effects of varying the annealing temperature in the amplification protocol. The F13A01, FESFPS, vWA Multiplex was used to amplify three DNA samples using the recommended 60°C annealing temperature or with an annealing temperature two degrees above or two degrees below the recommended temperature. Lanes labeled 1, 2, 3 contain the amplified DNA samples; lanes labeled N contain the negative control amplification (i.e., no template DNA added); lanes labeled L contain a mixture of the allelic ladders for the multiplex.

Figure 7. Amplification of genomic DNA samples using the SilverSTR™ III Multiplex. The twenty-one samples in the numbered lanes have each been amplified simultaneously at the loci D16S539, D7S820, and D13S317. The amplified fragments were separated in a 4% denaturing polyacrylamide gel and detected by silver staining.

Figure 8. Amplification sensitivity using the SilverSTR™ III Multiplex. K562 DNA ranging from 25ng to 0.5ng was amplified using the SilverSTR™ III Multiplex, separated in a 4% denaturing polyacrylamide gel, and detected by silver staining. Lanes 1-6 contain amplified K562 DNA using 25, 10, 5, 2, 1, and 0.5ng template DNA, respectively; lane 7 contains a negative control amplification reaction (i.e., no template DNA added).

Figure 9. Effects of varying the annealing temperature in the amplification protocol. The SilverSTR™ III Multiplex was used to amplify three DNA samples using the recommended 60°C annealing temperature or with an annealing temperature two degrees above or two degrees below the recommended temperature. Lanes labeled 1, 2, 3 contain the amplified DNA samples.