Home About Us Laboratory Services Forensic Science Communications Back Issues April 2003 Casework: A Y-STR Triplex for Use After Autosomal...

Casework: A Y-STR Triplex for Use After Autosomal Multiplexes, Chang Yuet Meng, Forensic Science Communications, April 2003

Casework: A Y-STR Triplex for Use After Autosomal Multiplexes, Chang Yuet Meng, Forensic Science Communications, April 2003


April 2003 - Volume 5 - Number 2

Research and Technology

Casework: A Y-STR Triplex for Use After Autosomal Multiplexes

Chang Yuet Meng
Graduate Student
School of Biological Sciences
Flinders University of South Australia
Bedford Park, Australia

Katrin Both
Senior Forensic Scientist
Forensic Science Centre
Adelaide, Australia

Leigh A. Burgoyne
School of Biological Sciences
Flinders University of South Australia
Bedford Park, Australia

Abstract | Introduction | Materials | Methods | Results
Discussion | Conclusion | References


The increase in the evidential utility of Y-STR multiplexes does not rise with loci number the same way as it does with unlinked autosomal loci. This might encourage consideration for using small Y-STR multiplexes. In this study, one practical Y-STR triplex that consists of DYS438, DYS390, and DYS439 was used to analyze 13 evidentiary stains from two sexual assault cases after using the AmpFlSTRâ Profiler PlusTM (PE Biosystems, Foster City, California) multiplex. In Case 1, the Y-STR profiles established the minimum number of male contributors in four out of five multiple-source stains and confirmed the identity of the suspects who contributed to the mixed autosomal profiles. In Case 2, the Profiler PlusTM results of the three stains showed two mixed male/female profiles and one stain with no result. A Y-profile consistent with the suspect’s profile was obtained in all three stains. The ability of the Y-triplex to selectively amplify the limited number of male cells in the third stain demonstrates the sensitivity of the Y-system. The results indicated that with mixed stains, the use of a small Y-STR multiplex in addition to autosomal STRs for forensic casework can help to discriminate multiple suspects and is more useful than using only autosomal STRs.


DNA mixtures from males and females are frequently encountered in forensic casework, particularly in rape and sexual assault cases. Often, these mixed stains are in limited quantity or of poor quality, thus they pose a challenge in the resolution and interpretation of their resulting genetic profiles. Complementing existing STR analysis with validated Y-STR systems appears to be the current as well as future approach in forensic casework involving mixtures of body fluids. Besides their ability to track paternal lineages, Y-STRs are important forensic tools in mixture analysis largely because of their ability to target and detect male-specific DNA (Jobling et al. 1997; Kayser et al. 1997; Kloosterman 1999). The use of Y-multiplexes to detect semen traces arising from multiple sources could obviate problems in the analysis and interpretation of complex autosomal profiles. However, the optimal type, number, and size of loci to be included in Y-STR multiplexes for practical forensic use are not yet clear.

Y-STR haplotyping in forensic laboratories worldwide is reaching for more loci but is currently centered on a nine-locus core set mentioned in the Y-STR Haplotype Reference Database (http://www.ystr.org). The National Institute of Standards and Technology is developing high-performance Y-STR multiplexes that combine more than 20 loci (http://www.cstl.nist.gov/biotech/strbase/). However, unlike the autosomal loci, a high number of Y-STR loci in multiplexes may be of little evidential value. Theoretically, as the number of loci in a Y-multiplex increases, the confidence in identifying a Y-chromosome and the ability to exclude increases, but incremental gains in exclusion power decreases. Additionally, as multiplexes get more loci, the compromises in amplification conditions usually lower the sensitivity. Thus, a small triplex may be the optimum loci size to use in a Y-STR multiplex. Whereas new, larger multiplexes such as Y-megaplexes certainly have advantages in studying the evolution of human Y-chromosomes, small Y-multiplexes consisting of three to five loci might be much more practical and robust. They may be more sensitive to low template levels and sufficient in discriminatory power for forensic application if used as an additional system to clarify and confirm autosomal DNA profiles.

This study investigates the application of one small Y-multiplex to see what advantages it may have when used in conjunction with a common autosomal multiplex such as the Profiler PlusTM. Thirteen evidentiary stains in two sexual assault cases were used. The Y-triplex consists of one locus from the core set of Y-STRs, DYS390 (Y-STR Haplotype Reference Database), and two relatively new Y-STR loci, DYS438 and DYS439 (Ayub et al. 2000). DYS390 was chosen because it has complex mutational features (Forster et al. 1998). DYS438 was chosen because it is a pentanucleotide repeat locus and is expected to produce minimal stutter artifact in the PCR. DYS439 has been found to exhibit high locus diversity values in populations with similar ancestry, like the Malaysian ethnic groups (Ayub et al. 2000; Hou et al. 2001).


Case 1

A group of men entered a house to rob the occupants. After taking the valuables, some of the men raped the female family member. Ten stains were recovered from the crime scene (M1, M2, M3, M4, M5, M6, M7, M8, M9, and M10) and were submitted for DNA analysis. Reference bloods were collected from the six suspects involved (S1, S2, S3, S4, S5, and S6). No vaginal swab from the victim was submitted for analysis.

Case 2

A naked woman was found dead in a drain. The case was classified as murder and a suspect was arrested. Police recovered two stains from the interior of the suspect’s car—one from the floor mat (N1) and one from the seat cushion (N2). A third stain was taken from a clothing item found inside the car (N3). The three stains were submitted for DNA analysis.


For both cases, genomic DNA was isolated from the reference bloods of suspects and victims using the Chelex method (Walsh et al. 1991). Because the ten evidentiary stains in Case 1 were semen traces, they were extracted using a slightly modified differential lysis procedure. The fabric materials (~ 0.5cm X 1cm) containing the stains were incubated for one hour with 700µl of sterile water with frequent vortexing. The fabric was placed in a spin basket, and the tube was spun down for five minutes to pellet crude DNA. The samples were then incubated at 56°C with 5µl of proteinase K (10mg/ml) and 100µl of sterile water for 30 minutes. After spinning down, the supernatant containing the epithelial fraction was isolated. The sperm pellet was washed twice with 250µl of sperm wash buffer (10mM Tris-HCl pH 7.5, 10mM EDTA, 50mM NaCl, 2% SDS) and spun down for ten minutes after each wash to collect the sperm fraction. The epithelial fraction was digested with 30µl of 20% Chelex. The sperm fraction was digested with 100µl of 5% Chelex, 13.5µl Proteinese K, and 10µl of DTT. The sperm fraction was then spun down twice to collect the supernatant, then concentrated to a volume of 40µl using MicroconTM B100 microconcentrators (Amicon, Incorporated, Beverly, Massachusetts) according to the manufacturer’s instructions.

For the three stains in Case 2, two procedures were performed—a differential procedure (as above) and a nondifferential procedure. In the nondifferential extraction, the three stains were extracted without the female epithelial digestion. The stains were saturated directly with 100µl 5% Chelex, 13.5µl of proteinese K, and 10µl of DTT, then incubated 56°C for one hour with the subsequent protocols performed as in Case 1. However, quantitation of the extracts from both cases was not performed. In this study’s validation work on the Y-triplex, the suitable target level of DNA for this amplification system was between 0.3-1ng of template DNA (data not shown). The sensitivity of the Y-triplex was determined as 0.1ng.

Autosomal STR Analysis

The AmpFlSTRâ Profiler PlusTM kit was used to perform autosomal STR profiling according to the manufacturer’s instructions. Between 10-15µl of the extracts were used for amplification on the GeneAmpâ PCR System 9600 (PE Applied Biosystems, Foster City, California). The PCR products were analyzed on the ABI PRISMTM 377 DNA Sequencer (PE Applied Biosystems, Foster City, California) using the internal size standard GSâ-400HD ROX (PE Applied Biosystems, Foster City, California). Resulting fragments were sized with GeneScanTM (PE Applied Biosystems, Foster City, California) using 75 RFU threshold for making allele calls in GenotyperTM (PE Applied Biosystems, Foster City, California). A threshold of 35 RFU was used for NR allele.

Y-STR Analysis

The three Y-STR primers from Geneset Oligos (Lismore, Australia) were labeled with fluorescent dyes (DYS438 with 6-FAM, DYS390 with JOE, and DYS439 with TAMRA). The PCR reaction mix of 15µl contained DYS438 primer pair (each at 0.13µM), DYS390 primer pair (each at 0.10µM), DYS439 primer pair (each at 0.50µM), 1X GeneAmpâ PCR Buffer II (PE Applied Biosystems, Foster City, California) containing 10mM Tris-HCl pH 8.3, 50mM KCl, 1.5mM MgCl2, 200µM of each dNTPs (PE GeneAmpâ), 1.2U AmpliTaq GoldTM polymerase (PE Applied Biosystems, Foster City, California), and 16mg/ml BSA (Sigma Chemical Company, St. Louis, Missouri). Amplification was carried out on the GeneAmpâ PCR System 9600 with the cycling conditions at 95°C for 11 minutes, followed by 30 cycles of 94°C for 1 minute; 57°C for 3 minutes; 72°C for 3 minutes, and ending with 72°C for 20 minutes. Between 4-7µl of the extracts were used in the PCR assay. Y-STR alleles were designated by the variable number of repeat units at the respective locus according to recommendations by the International Society of Forensic Haemogenetics. Allele designation was achieved by using two Y-STR control samples with known repeat numbers, supplied by M.A. Jobling, University of Leicester, United Kingdom (personal communication). The PCR products were analyzed on the 377 DNA Sequencer using the same internal size standard and 35 RFU threshold for allele sizing in GeneScanTM.


The Profiler PlusTM results for the sperm fractions of the ten stains in Case 1 are presented in Table 1. There are five multiple-source stains (M1, M2, M4, M6, and M7), four single-source stains (M3, M5, M8, and M9), and one stain with no profile (M10). Interpretation of the five mixed profiles is limited by partial profiles and shared alleles of some of the suspects.

Figure 1A shows three horizontal line graphs one below another, each with vertical spikes in various places along the line and square identification markers to each spike depicting the Profiler PlusTM results.  Figure 1B shows a horizontal line graph with three major vertical spikes depicting the Y-triplex results in GeneScanTM.  Figure 1C shows a line graph similar to B with the results in GeneScanTM.

Figure 1 Electropherograms of stain M2 in Case 1.
A Profiler PlusTM results in GenotyperTM. Based on the five alleles detected at D8 locus, there are at least three contributors to this complex DNA profile. Suspects S4, S5, and S6 cannot be excluded from contributing to this mixture.
Y-triplex results in GeneScanTM. The Y-profile confirms the three male contributors to the complex stain M2.
C DYS439 system in GeneScanTM.
Click to enlarge image.

Ten exclusions were drawn based on just the Profiler PlusTM results:

  • S4, S5, and S6 could not be excluded from contributing to stain M2
  • There were five alleles at D8 that are not stutter bands based on their peak heights (Figure 1A).
  • S4 and S6 could not be excluded as the main contributors to stain M4
  • There were five alleles at VWA where one was an NR allele (VWA*18= 40 RFU). The presence of a very minor third contributor could not be established.
  • Both S4 and S5 could not be excluded from contributing to stain M7.
  • Stain M1 could not exclude S6 and showed a partial profile of S4.
  • S5 could not be excluded from stains M3 and M8.
  • S1 could not be excluded from stains M5 and M9.
  • Stain M6 only showed a partial profile of S1 and S4.
  • Stain M10 did not yield any profile.

The Y-triplex results (Table 2) confirmed that stains M2 and M4 have at least three contributors, whereas stains M1 and M7 have at least two contributors. Stains M3, M5, M8, and M9 have only one male contributor. The Y-triplex was optimized and tested on 96 female samples. The three loci were found to be male-specific with no Y-STR alleles detected (data not shown). Allele and haplotype frequencies of the three Y-STRs were determined in 338 unrelated individuals in the Malaysian population, and the three loci were found to show high locus diversity values. Additionally, DYS438 was found to show a strong ethnic affiliation between the Caucasian and the non-Caucasian groups studied (Chang et al. 2002).

The six suspects in Case 1 are Malays and have the same unique DYS438-allele 10, common in the Malaysian Malay population group (data not shown). DYS390 alone provided a firm exclusion of two of the suspects, S2 and S3, whose DYS390 alleles were not seen in any of the stain profiles (Table 2). DYS439 was the most discriminative among the three loci. The presence of three different DYS439 alleles in stain M2 confirmed that there were at least three male contributors to this stain, whereas the presence of two different DYS439 alleles in stain M7 confirmed that there were at least two male contributors to this stain. Based on the Y-profile alone, having the same haplotype, S1 and S4 could not be differentiated. The frequency of this shared haplotype in the Malaysian Malay population is 0.0885 (data not shown). However, both suspects could be differentiated based on their autosomal profiles. This indicates that both multiplexes when combined give a uniquely powerful and discriminatory DNA analysis that cannot be achieved using either multiplex alone.

Figure 2A shows three horizontal line graphs one below another, each with vertical spikes in various places along the line and square identification markers to each spike depicting the Profiler PlusTM results.   Figure 2B shows a series of three horizontal line graphs with three major vertical spikes and a broad peak at the far right depicting the Y-triplex results in GeneScanTM.

Figure 2 Electropherograms of the stains in Case 2.
A. Profiler PlusTM results (in GenotyperTM) of stain N1 in the suspect’s car. The DNA profile is a mixture in which the major component is from the victim. The suspect is not excluded as the source of the minor component (detailed results in Table 4).
B. Y-triplex results in GeneScanTM of the three stains: N1, N2, and N3. Each Y-profile depicts three alleles: DY390 is on the left, DYS438 is in the middle, and DYS439 is on the right.
Note that the RFU in the third stain is very weak, and the broad peak at the far right is the dye artifact from TAMRA in the DYS439 system. This stain failed to give any autosomal profile with the Profiler PlusTM multiplex.
Click to enlarge image.

In Case 2, using the differential extraction protocol, stains N1, N2, and N3 gave no autosomal STR or
Y-STR results. This could not be attributed to the inefficiency of the differential extraction protocol because the same procedure gave positive results in all but Case 1, stain M10. However, using a nondifferential extraction protocol followed by a Microcon step, a mixed male/female autosomal profile was observed in both stains N1 and N2, whereas stain N3 still failed to yield any profile
(Table 3). A large imbalance between the X-allele and Y-allele at the amelogenin locus was observed in the two mixed profiles, confirming the presence of only a small amount of the perpetrator’s DNA (Figure 2A). The suspect and victim could not be excluded from contributing to both stains N1 and N2 based on their autosomal profiles. It should also be noted that the Profiler PlusTM process resulted in the two stains N1 and N2 showing the presence of an another allele (D3*16), which was present in neither the suspect’s nor the victim’s profile.

The three nondifferential extracts amplified with the Y-triplex yielded a Y-profile consistent with the suspect’s profile (Table 4). Even the third stain, which failed to yield an autosomal profile, produced a weak Y-profile (Figure 2B). The frequency of finding the haplotype in stains N1 and N2 in the Malaysian population is 0.0088 (data not shown). Although the victim’s profile is the major component in the two stains N1 and N2, the presence of her blood in the suspect’s car should be probative even without his alleles being present. However, because the stains were retrieved from the suspect’s car, one can argue that there is a possibility of detecting a trace amount of his own DNA in his car. The fact that no sperm cell was visibly detected from the Christmas Tree Staining Test and that the results from the differential extraction protocol were negative, the Y-STR results obtained from stains N1 and N2 could have originated from underlying traces of his epithelial, saliva, or other cells. It is possible that the male DNA contribution was unrelated to the crime under investigation; therefore, the DNA result could not conclude that the suspect had committed the crime, but only that the victim had bled in his car.


The ability to detect a male profile in stain N3 clearly indicates the sensitivity of the Y-triplex and demonstrates the usefulness of this simple Y-system in analyzing stains that are too weak to give autosomal profiles. For a variety of technical reasons, the low number of loci in a triplex potentially enables a higher sensitivity compared to a larger multiplex such as the Profiler PlusTM system. Four significant reasons for the higher sensitivity of the Y-triplex could be:

  • The cycling conditions for each locus is never the same because the match of more than one is always a compromise that becomes more significant as more loci are included in the compromise.
  • The choice of exact sequence for the primers loses degrees of freedom as the primer numbers increase because of the need to avoid inadvertent cross-homology with resultant consumption of the PCR amplification by cross-locus primer-dimers. This progressive restriction of freedom on the choice of primer sequences as multiplexes increase in size reduces the freedom to optimize the overall amplification.
  • Factors 1 and 2 combined largely determine the number of productive PCR cycles possible. The number of PCR cycles is 30 for the Y-triplex and 28 for the Profiler PlusTM.
  • The ratio of primer to template DNA is another possible reason. With the same amount of total DNA, the Y-linked primers have multiples of one copy at the haploid locus to bind to, whereas the autosomal primers have the same multiples of two copies at a heterozygous locus upon which to bind.

The higher sensitivity of Y-STR analysis compared to autosomal STR analysis was observed by Betz et al. (2001) when a male profile was successfully detected from the vaginal epithelial cells in a rape case that failed to yield any autosomal profile. A number of studies on the application and validation of small Y-multiplexes with only three to six loci have been reported in forensic casework (Corash et al. 2001; Dekairelle and Hoste 2001; Prinz et al. 1999). This author’s experience suggests that Y-multiplexes should be constructed from a choice of highly informative Y-STRs optimized for a particular population group. Y-STRs with high locus diversity values in populations from similar ancestry should be chosen.

The analysis and interpretation of mixtures from autosomal STRs have generally relied on the observation of the number of alleles at each locus and the relative allele peak heights or peak areas (Clayton et al. 1998; Evett et al. 1990; Gill et al. 1998). The degree of contribution by multiple suspects in a mixed profile is usually based on the ratio of the allele peak heights or peak areas. However, this estimation could become difficult for very weak mixed profiles when there is interference from stutter and artifact peaks. For stain M2 in Case 1, the peak heights of the DYS439 alleles in the Y-profile directly indicated the contribution of the three suspects to the mixture (Figure 1C). It should be noted that a comparison of the peak height and peak area between the two systems might provide useful information in assisting autosomal mixture interpretation. To explore whether the Y-STR estimations relate directly to autosomal peak heights will take further testing on more casework mixtures. With no size overlap between DYS438 and DYS439 alleles, the peak height balance of the triplex could be improved by labeling DYS439 with 6-FAM. This will also eliminate the presence of dye artifacts from the TAMRA system. However, in the Y-triplex, these dye artifacts do not fall within the allelic range of the three loci (Figure 2B).

Although mixed stains from only two cases were tested, the Y-STR results indicate that using a small Y-multiplex is practical in a forensic environment. It should also be noted that with the Y-chromosome, the discriminatory power did not translate into evidential values as it did with the autosomes. Because the three chosen loci have alleles sized between 190-260bp, they were ideal for analyzing DNA with environmentally challenged quality and quantity.


The Y-STR triplex is useful to establish the minimum number of male contributors in multiple-source stains and to confirm the multiple suspects who contributed to complex DNA profiles. The system is also able to detect and characterize very limited numbers of male cells in a stain that is too weak to yield any autosomal result. Perhaps the most important use of the system is the ability to exclude individuals from complex mixtures. Both the Y-linked loci DYS390 and DYS439 are suitable markers to add to existing STRs in forensic casework. It is suggested that another Y-STR that is more polymorphic than DYS438 could be included to enhance the discrimination power of this Y-system.


Ayub, A., Mazhar, K., Mehdi, S. Q., Mohyuddin, A., Qamar, R., Tyler-Smith, C., and Zerjal, T. Identification and characterization of novel human Y-chromosomal microsatellites from sequence database information, Nucleic Acids Research (2000) 28(2):e8, i-v.

Betz, A., Bäßler, G., Dietl, G., Pflug, W., Steil, X., and Weyermann, G. DYS STR analysis with epithelial cells in a rape case, Forensic Science International (2001) 118:126-130.

Chang, Y. M., Both, K., and Burgoyne, L. A. Genetic variation at three DYS-STR loci and amelogenin between the Australian Caucasians and three Asian ethnic groups, Journal of Forensic Sciences (2002) 47(4):918-921.

Clayton, T. M., Gill, P., Sparkes, R., and Whitaker, J. P. Analysis and interpretation of mixed stains using DNA STR profiling, Forensic Science International (1998) 91:55-70.

Corash, D., Filgueira Risso, L., Marino, M., Penacino, G., and Sala, A. Routine Y-STR typing in forensic casework, Forensic Science International (2001) 118:131-135.

Dekairelle, A. F. and Hoste, B. Application of a Y-STR pentaplex PCR (DYS19, DYS389 I, and II, DYS390 and DYS393), Forensic Science International (2001) 118:122-125.

DNA recommendations: Further report of the DNA Commission of the International Society for Forensic Haemogenetics regarding PCR-based polymorphisms in STR (short tandem repeats), International Journal of Legal Medicine (1994) 107:159-160.

Evett, I. W., Buffery, C., Stoney, D. A., and Willot, G. A guide to interpreting single locus profiles of DNA mixtures in forensic cases, Journal of Forensic Sciences (1990) 31(1):41-47.

Forster, P., Benkmann, H., Brinkmann, B., Kayser, M., Meyer, E., Pfeiffer, H., and Roewer, L. Phylogenetic resolution of complex mutational features at Y-STR DYS390 in Aboriginal Australians and Papuans, Molecular Biology and Evolution (1998) 15(9):1108-1114.

Gill, P., Buckleton, J., Clayton, T., Pinchin, R., Sparkes, R., and Whitaker, J. Interpreting simple STR mixtures using allele peak areas, Forensic Science International (1998) 91:41-53.

Hou, Y. P., Li, Y. B., Prinz, M., Wu, J., Zhang, J., and Zhang, S. Z. Allele sequences of six new Y-STR loci and haplotypes in the Chinese Han population, Forensic Science International (2001) 118:147-152.

Jobling, M. A. University of Leicester, United Kingdom, personal communication, July 31, 2001.

Jobling, M. A., Pandya, A., and Tyler-Smith, C. The Y chromosome in forensic analysis and paternity testing, International Journal of Legal Medicine (1997) 110:118-124.

Kayser, M., de Knijff, P., Dieltjes, P., Krawczak, M., Nagy, M., Pandya, A., Roewer, L., Tyler-Smith, C., and Zerjal, T. Applications of microsatellite-based Y chromosome haplotyping, Electrophoresis (1997) 18:1602-1607.

Kloosterman, A. Application of Y-chromosome specific STR typing in forensic stains, First International Conference on Forensic Human Identification in the Millennium (1999). Available: http://www.forensic.gov.uk/forensic/conference/papers/y_chromosome/html

Prinz, M., Baum, H., Ishii, A., Sansone, M., and Shaler, R. Experiences with using Y chromosome specific STRs in forensic casework. In: Tenth International Symposium on Human Identification, Orlando, Florida, 1999.

Short Tandem Repeat DNA Internet Database. National Institute of Standards and Technology, Gaithersburg, Maryland. Available: http://www.cstl.nist.gov/biotech/strbase

Walsh, P. S., Higuchi, R., and Metzger, D. A. Chelexâ 100 as a medium for simple extraction of DNA for PCR-based typing from forensic material, Biotechniques (1991) 10(4):506-513.

Y-STR Haplotype Reference Database. Institute of Legal Medicine, Humboldt-Universitat Berlin, Germany. Available: http://www.ystr.org

Top of the page