Alternative direct‐to‐amplification cell lysis techniques for forensically relevant non‐sperm cells

While efforts have been made to reduce the pervasive backlog of sexual assault evidence collection kits, the actual laboratory process remains very time‐consuming due to the requirement of a differential lysis step before DNA purification, as well as intricate mixture analysis towards the end of the DNA workflow. Recently, an alternative, direct‐to‐amplification sperm lysis method (using 1 M NaOH) was identified. However, a direct cell lysis method for non‐sperm cells has not been identified yet. Thus, the primary objective of this work was to find an alternative method that is quick, inexpensive, and does not require multiple purification steps for the lysis of non‐sperm cells in sexual assault samples. In this study, vaginal swab samples were lysed with the control method, prepGEM™, as well as six alternative reagents: alkaline buffer with 25–200 mM NaOH, high‐salt stain extraction buffer, modified radioimmunoprecipitation assay (RIPA) buffer, mammalian protein extraction reagent (M‐PER™), digitonin buffer, and urea/thiourea buffer. Quantification using Quantifiler® Trio of vaginal and semen lysates revealed that the alkaline (25 mM NaOH) and M‐PER™ methods were efficient for the lysis of vaginal epithelial cells without substantial sperm cell lysis. Following quantification, analysis of STR profiles from vaginal lysates revealed that the M‐PER™ method showed promising results across all metrics examined, including the percentage of detected STR alleles, mean peak heights, peak height ratio, and interlocus balance. Thus, this method was recommended as an alternative to the traditional differential lysis method for non‐sperm cells given its ability to produce amplification‐ready lysates without any DNA purification step.


| INTRODUC TI ON
Deoxyribonucleic acid (DNA) testing is a valuable tool recognized by law enforcement agencies for criminal investigations, including sexual assaults.Sexual assault is a prevalent crime in today's society, with an average of 463,634 victims (age 12 or older) in the United States annually [1].Due to the frequent number of sexual assaults, laboratories receive large amounts of evidence for processing, but many do not have the resources to invest in new technology or hire the additional employees needed to examine all of the cases quickly [2].As a result, there is a delay in processing sexual assault evidence samples; it can take months or even years for a laboratory to complete a case, leading to a sustained backlog [3].Despite recent attempts to resolve the issue of untested sexual assault evidence kits (SAEKs) in the United States, the problem persists due to new laws in several states mandating the submission and testing of all collected SAEKs.In addition, although violent crime has decreased since 2018, the latest data from the Bureau of Justice Statistics reveals that sexual assault and rape still constitute around 22% of reported victimizations to the police [4].
A commonly collected sample after a sexual assault is an intimate swab from the vagina of the female victim, which may contain mixtures of spermatozoa from the male perpetrator and vaginal mucosal epithelial cells from the victim.Therefore, the most common laboratory approach for analyzing sexual assault evidence is to separate sperm cells from non-sperm cells in order to isolate the DNA of a suspected male offender from the DNA of a victim [5].Although differential lysis is the most common technique for processing such samples, this procedure involves extensive incubation times, multiple purification steps, and repeated manual tube-to-tube transfers (which causes an increased risk of error and/or cross-contamination)-ultimately hindering the ability of a laboratory to keep pace with submissions [6][7][8][9].In fact, the separation of the female and male fractions can take up to eight hours, which contributes significantly to the existing backlog.Furthermore, the separation process itself can lead to the loss of 60%-90% of the male DNA, which is often the most crucial component to retain [10][11][12][13].Thus, an approach to mitigate this issue is to develop an alternative cell lysis method that would decrease the processing time, limit tube-to-tube transfers, minimize sample loss, and perform comparably, if not better for comprehensively lysing the cells that are observed in sexual assault samples.
In a recent study by Schellhammer et al., alternative direct-toamplification sperm cell lysis methods were explored in order to provide a faster and cheaper alternative to the conventional sperm cell lysis technique [14].Of the methods tested, the alkaline lysis method-which uses 1 M sodium hydroxide (NaOH) for five minutes at 75°C-was ultimately selected as the best alternative approach to accomplish sperm cell lysis [14].However, identification of a similar direct-to-amplification lysis method for the non-sperm fraction is also needed, as this is critical to achieve successful centrifugal separation of the cellular components.One alternative approach for direct cell lysis of non-sperm cells includes the use of the prepGEM™ Universal kit (prepGEM; microGEM™, Charlottesville, VA).This kit can be used for extracting DNA from saliva, blood, and tissues by utilizing the thermostable proteinase EA1, which operates only at 75°C, to lyse cells and degrade proteins [15].Not only is this a quick and simplified method that provides qPCR and STR-ready DNA in 15 min, but it is also performed in a single tube, reducing tube-totube transfers and thus minimizing the potential for contamination [15].However, this kit is relatively expensive and has been reported to generate low-level profiles, allelic drop-out, and inhibition [16].
Therefore, alternative non-sperm lysis techniques must be explored, including those that are cheaper and could maximize non-sperm lysis efficiency while preventing or reducing amplification inhibition.This alternative cell lysis technique for non-sperm cells, when coupled with alkaline lysis (1 M NaOH) for direct-to-amplification sperm cell lysis, would enable an alternative in-tube differential technique that would be quick, inexpensive, and obviate the need for multiple purification steps [17,18].Such alternative non-sperm cell lysis techniques could involve the use of a variety of chemical lysis buffers.Alkaline extraction is an inexpensive, straightforward, quick, and effective method for the solubilization of proteins (such as membrane proteins).In addition, the fundamental structure of DNA is comparatively stable in an alkaline solution [19][20][21].However, the alkaline approach can also lead to lysis of sperm cells if used at high concentrations (e.g., 1 M NaOH) [14,19,22].Thus, to differentially lyse epithelial cells and not sperm cells, lower concentrations of NaOH would need to be utilized [14].Radioimmunoprecipitation assay (RIPA) buffer and urea/thiourea lysis buffer are also commonly used for rapid and efficient cell lysis, as well as solubilization of proteins, from mammalian cells-releasing cytoplasmic, nuclear, and mitochondrial proteins [23][24][25][26][27][28].Furthermore, a simple high salt "stain extraction buffer" composed of 300 mM NaCl, 10% glycerol, 1 mM EDTA, and 25 mM Tris-HCl (pH 8.0) has been described in previous studies as a way to aid in the differential lysis of non-sperm mammalian cells [29,30].This buffer has been used for the lysis of sf9 cells, which are a cell line that exhibits epithelial morphology [29,30].Finally, digitonin, a non-ionic detergent that is derived from the purple foxglove plant Digitalis purpurea, may potentially be helpful for the differential lysis of specific mammalian cells [31].Digitonin is a weak detergent • Alternative chemical methods for direct-to-amplification lysis of non-sperm cells were evaluated.
• Alkaline (25 mM NaOH) and M-PER™ methods lysed non-sperm cells without substantial sperm cell lysis.
• M-PER™ lysis method can efficiently provide high-quality suitable downstream STR profiles.that permeabilizes the plasma membrane causing it to release cytosolic components, while leaving other organelles intact [32].Similarly, Thermo Fisher Scientific recently developed the M-PER™ Mammalian Protein Extraction Reagent, which allows for the extraction of proteins from the cytoplasm and nucleus of mammalian cells.This reagent is gentle, as it uses a mild and non-denaturing detergent for breaking down cell membranes [23,33,34].
By using one or more of these alternative techniques for lysis of the non-sperm fraction of sexual assault samples, it may be possible to liberate the DNA in a simple and quick manner, without the need for multiple tube-to-tube transfers or formal purification.Thus, several alternative methods for differential cell lysis of non-sperm cells (e.g., immature germ cells, epithelial cells, leukocytes) were evaluated and compared in an effort to identify the best performing methods that would be most amenable to the current forensic DNA workflow.

| Sample collection and preparation
For this study, one vaginal swab and one semen sample were collected from five female and male volunteer donors, respectively according to the Virginia Commonwealth University (VCU) approved Institutional Review Board (IRB) protocol HM20002931.Semen was diluted 1:10 by volume using Gibco™ 1X Dulbecco's phosphate-buffered saline (DPBS) (Fisher Scientific; Waltham, MA).Cells were eluted from vaginal swabs by submerging the whole swab cutting in 400 μL DPBS and incubating at 37°C for two hours, with brief vortexing every 15 min.

| prepGEM™ universal kit (non-sperm cell lysis control)
For this method, ten microliters of either the vaginal sample or the 1:10 semen sample were combined with 0.5 μL prepGEM™ enzyme, 5.0 μL 10X Blue buffer, and 34.5 μL HyPure Molecular Biology Grade Water (MBG H 2 O; GE Healthcare Life Sciences; Chicago, USA) in 0.2 mL PCR tube strips (Thermo Fisher Scientific).Samples were then incubated using the ProFlex™ 3 × 32-well PCR System (Applied Biosystems™, Waltham, MA) as follows: 75°C for five minutes, then 95°C for two minutes.Lysates were stored at −20°C until further processing.

| Alkaline lysis (1 M NaOH) (sperm cell lysis control)
Alkaline lysis (1 M NaOH) was used as the control for semen samples as it leads to complete lysis of sperm cells [14].For this method, ten microliters of 1:10 semen was incubated in 7 μL of DPBS and 4 μL of 1 M NaOH (Thermo Fisher Scientific) at 75°C for five minutes in 0.2 mL PCR tube strips using the ProFlex™ 3 × 32-well PCR System.Succeeding incubation, 4 μL of 1 M Tris-HCl (pH 8.0; Invitrogen; Carlsbad, CA) was added to the samples, which were then briefly vortexed [14].Lysates were stored at −20°C until further processing.Lysates were stored at −20°C until further processing.

| High salt stain extraction buffer
High salt stain extraction buffer was prepared by adding 25 mM Tris-HCl (pH 8.0), 300 mM NaCl, 10% glycerol (Mallinckrodt; Raleigh, NC), and MBG H 2 O.To achieve cell lysis, ten microliters of the vaginal eluate was combined with 15 μL of the buffer in a 0.2 mL PCR tube strip and incubated at room temperature for 15 min.The lysates were stored at −20°C until further processing.

| Modified RIPA lysis buffer
Two modified RIPA lysis buffers were prepared for testing as previously described [27].However, each buffer was prepared without the recommended sodium dodecyl sulfate (SDS), and one included additional proteinase K (proK) (Thermo Fisher Scientific).
The modified RIPA lysis buffer contained 150 mM NaCl (Thermo Fisher Scientific), 50 mM Tris-HCl (pH 7.5), 1% Triton X-100 (Thermo Fisher Scientific), 0.5% sodium deoxycholate (Sigma Aldrich; St. Louis, MO), and MBG H 2 O.For the modified RIPA buffer lysis with proK, 0.72 mg/mL proK (final concentration) was added to the above recipe.For this method, ten microliters of the vaginal eluate and 15 μL of the modified buffer were added to a 1.7 mL microcentrifuge tube (Thermo Fisher Scientific) and placed in a thermomixer (Eppendorf; Hamburg, Germany) at 56°C for 30 min at 1500 rpm.Lysates were stored at −20°C until further processing.

| M-PER™ mammalian protein extraction reagent
Either ten microliters of the vaginal eluate or 1:10 semen was combined with 15 μL of the M-PER™ Mammalian Protein Extraction Reagent (Thermo Fisher Scientific), and the 1.7 mL microcentrifuge tube was placed in a thermomixer at 56°C for 30 min at 1500 rpm.
Lysates were stored at −20°C until further processing.

| Digitonin buffer
For this method, ten microliters of vaginal eluate were submerged in 15 μL of the buffer solution in a 1.7 mL microcentrifuge tube, which included 50 mM HEPES-NaOH (HEPES: Sigma Aldrich), 150 mM NaCl, 10% Glycerol, and 1% digitonin (Sigma Aldrich).The samples were placed in a thermomixer at 56°C for 30 min at 1500 rpm.
Lysates were stored at −20°C until further processing.

| Microscopy
Microscopy was used as a preliminary screening method to check the lysis efficiency for vaginal epithelial cells (from vaginal samples) and sperm cells (from semen samples).Prior to and following cell lysis, Kernechtrot Picroindigocarmine staining (KPICS) was performed on each sample in order to visually gauge the effectiveness of each lysis method.For this, a total of five microliters of either vaginal or semen sample were spotted onto a microscope slide, allowed to dry at 56°C, fixed with Sprayfix® Cytology Fixative (Leica Biosystems; Wetzlar, Germany), and dried for five minutes at room temperature.Next, each slide was stained with one drop of Kernechtrot stain (Serological Research Institute; Richmond, CA) for 15 min, rinsed with water, stained with one drop of Picroindigocarmine stain (Serological Research Institute) for 20 s, and then gently rinsed with water.After staining, 15 μL of MBG H 2 O was added to the slide and a cover slip was placed on top.When vaginal cells were treated with the urea/thiourea buffer, a white precipitate was observed upon heating the slides to 56°C for KPICS staining.
Consequently, samples treated with urea/thiourea were re-stained using trypan blue.For this, ten microliters of 0.4% trypan blue (Invitrogen) were added to 15 μL of sample, mixed via pipetting for ten seconds, and then spotted onto a microscope slide.
Regardless of the staining method used, stained cells were visualized under a Micromaster microscope (Thermo Fisher Scientific) using 400X total magnification.Each slide was scored using a 0-4+ scale, where "0" meant no cells were observed in the field, "1+" indicated there was one cell seen in some fields, "2+" indicated 1-5 cells were seen in most fields, "3+" meant 5-10 cells were observed in most fields, and "4+" meant >10 cells were observed in all fields.Five different fields-of-view were scored for each sample and a mean value was calculated.Methods that successfully lysed vaginal cells were selected for continued testing using semen samples.

| DNA quantification
In order to determine the total amount of human DNA obtained from each sample after treatment with each lysis method, the resulting vaginal cell lysates were quantified using the Quanti-filer™ Trio DNA kit (Applied Biosystems) on the QuantStudio™ 6 Plex Real-Time PCR system (Applied Biosystems).Manufacturerrecommended protocols were used, but with half-volume reaction conditions [35].This included 5 μL of Reaction Mix, 4 μL of Primer Mix, and 2 μL of template DNA per sample.These recommendations were followed for the prepGEM™, alkaline lysis (all concentrations), and M-PER™ experimental groups.However, the reagents used in the high salt stain extraction buffer, modified RIPA buffer, urea/thiourea buffer, and digitonin buffer are known to inhibit PCR; thus, 5X AmpSolution™ (Promega; Madison, WI) was also added to each quantification reaction (as well as DNA standards) for samples lysed with these methods.For these samples, each reaction included 5 μL of Reaction Mix, 4 μL of Primer Mix, 2 μL of 5X Amp Solution, and 2 μL of template DNA.For all quantification runs, standards and no template control were quantified in duplicate.Thermal cycling conditions were as follows: 95°C for two minutes, followed by 40 cycles of 95°C for 9 s and 60°C for 30 s.
Results were analyzed using the QuantStudio™ Real-Time PCR Software v1.3 (Thermo Fisher Scientific).
Various metrics from the quantification data were assessed for each run, including the slope, Y-intercept, and R 2 of the standard curve, as well as the cycle threshold (C q ) of the internal PCR control (IPC), to assure that they were within expected manufacturer ranges [35].Furthermore, the component and amplification plots were evaluated.The IPC C q values were compared across all lysis methods, with the expected range of 27-31 per the manufacturer's guidelines [35].
Inhibition was noted if samples produced an IPC C q value higher than the range, indicating delayed amplification; samples that produced a value lower than the range were interpreted as inconclusive.The multicomponent plots were examined for any discrepancies from the expected fluorescent signals of the three amplification targets, which should remain consistent and constant for the first 15 to 20 cycles before exhibiting exponential amplification of the PCR product.The passive reference dye signal was also analyzed to identify any deviations from a uniform, stable curve throughout the entire assay [35].
Total DNA yields were calculated by multiplying the concentration obtained from the small autosomal target by the total sample volume.
Male DNA yields were calculated by multiplying the concentration obtained from the Y target by the total sample volume.The means and standard deviations for each experimental group were calculated and compared in order to determine which methods would proceed to downstream analyses.Samples that produced "undetermined" quantification results were excluded from the calculations.An ANOVA was performed using the IBM® SPSS Statistics software to compare the DNA yields of the control method (prepGEM™) to the experimental lysis methods (α = 0.05).If the ANOVA resulted in a significant difference, a Tukey HSD test was performed to identify where these significant differences were.Experimental lysis methods that resulted in passing quality metrics and DNA yields from vaginal samples were then tested on semen samples.Microscopy and DNA quantification of semen samples were performed as noted above.
In order to estimate the percentage of sperm cells lysed by each method, the total human DNA yields obtained from semen samples were compared to the values obtained from the same semen samples treated with the control sperm cell lysis method (1 M NaOH).This was calculated by dividing the total small autosomal DNA concentration obtained in an experimental group for one donor by the DNA concentration obtained in the control method for the same donor and then multiplying it by 100; these values were then averaged across all donors.Given the reported expected normal proportions of sperm (88%) and non-sperm cells (12%) in semen [16,[36][37][38], 80% of the total DNA within semen should theoretically stem from spermatozoa, while the remaining 20% should originate from non-sperm cells.Therefore, an ideal non-sperm cell lysis method would be the one that shows no more than 20% of the total available DNA in a semen sample.
The degradation index (DI) was calculated and analyzed for all tested samples by dividing the concentration of the small autosomal target by the concentration of the large autosomal target.The average of all values within an experimental group was then calculated to get the final DI for the method.For samples within this study that exhibited an acceptable IPC (C q 27-31), the DI values were interpreted in four categories, as previously described [39]: 0-1.5 being non-degraded, 1.5-4 being mildly degraded, 4-10 being degraded, and >10 being severely degraded.Samples with an IPC value outside the acceptable range were considered uninterpretable when the DI was less than 1 -meaning that it was not possible to determine if the sample was inhibited, degraded, or both.

| STR amplification
Following human DNA quantification, Powerplex® Fusion 6C (Promega) was used to amplify 0.75 ng of input DNA on the ProFlex™ PCR System.PCR amplification was performed according to manufacturer recommendations, with modifications for half-volume reactions [40].Thus, 5 μL sample (at 0.15 ng/μL), 2.5 μL PowerPlex® Fusion 5X Master Mix, 2.5 μL PowerPlex® Fusion 5X Primer Pair Mix, and 2.5 μL amplification-grade water were combined for each reaction.Samples with DNA concentrations less than 0.15 ng/μL were concentrated using the Thermo Fisher Scientific Savant DNA 120 SpeedVac concentrator to obtain the desired concentration.
The positive and negative controls were included in the amplification reaction and the thermal cycling conditions were as follows: 96°C for 1 min; 29 cycles of 96°C for 5 s and 60°C for 1 min; 60°C for 10 min; and a 4°C hold (forever).

| Capillary electrophoresis and data analysis
STR amplicons were separated and analyzed using capillary electrophoresis (CE) on the Applied Biosystems™ 3500 Genetic Analyzer using Data Collection software v4.Samples and allelic ladders (1 μL) were combined with 9.5 μL of Hi-Di™ Formamide (Thermo Fisher Scientific) and 0.5 μL of WEN ILS 500 (Promega™).Injection parameters followed manufacturer recommended settings, which included a 36 cm capillary array (Thermo Fisher Scientific), POP-4® polymer (Thermo Fisher Scientific), and a 1.2 kV 15 s injection.Resulting STR profiles were analyzed using GeneMapper™ ID-X software v.6.1 (Thermo Fisher Scientific) according to manufacturer recommended settings [41], with an analytical threshold of 150 Relative Fluorescence Unit (RFU) and a stochastic threshold of 300 RFU.
The overall quality of the STR profile was determined by the percentage of expected STR alleles present, peak heights, intralocus peak height ratios, and interlocus balance.The percentage of STR alleles present in the sample was calculated by taking the number of detected alleles and dividing it by the number of expected alleles (determined by prior genotyping of each volunteer's known reference sample) and then multiplying that by 100.
The mean peak height of all STR alleles observed (except those at Amelogenin, DYS391, DYS576, and DYS570 loci) was calculated for each experimental group; to account for homozygosity, the peak heights for homozygous alleles were halved to represent each of the assumed two copies of the allele at that locus.The intralocus peak height ratios (PHR) were calculated by dividing the peak height of an allele with a lower RFU value by the peak height of a sister allele with a higher RFU value.The interlocus balance was also determined using the coefficient of variation (CV) for the locus peak height: total peak height (LPH:TPH) ratios of each locus (except Amelogenin, DYS391, DYS576, and DYS570) in the DNA profile.By measuring interlocus balance, peak height variation for each locus was compared to the average peak height across the entire profile.The desired CV should be ≤0.35,indicating that the peak heights at a specific locus do not differ more than 35% from those at other loci in the DNA profile [42].

| Microscopic evaluation of lysates
Prior to cell lysis, vaginal eluates resulted in a mean score of 4+ showing numerous intact cells per field.Following lysis with the non-sperm lysis control method (prepGEM™), vaginal eluates showed complete cell lysis when examined microscopically (mean score of 0).Similarly, treatment of vaginal eluates with the alkaline lysis conditions resulted in complete cell lysis across all donors, regardless of the concentration used.Further, eluates subjected to the high salt stain extraction buffer, M-PER™, and digitonin lysis buffer also exhibited complete cell lysis across all donors.Interestingly, treatment of vaginal eluates with the modified RIPA lysis buffer resulted in the visualization of intact cells (mean score of 3+); however, when proK was added to the modified RIPA lysis buffer, no intact vaginal epithelial cells were observed.Finally, complete lysis of vaginal cells was also achieved when eluates were treated with the urea/thiourea reagent.Given the lysis success observed microscopically (data not shown), all experimental methods, except for the modified RIPA lysis buffer (without proK), were subjected to downstream testing.Ideally, an optimal lysis method for this study would be one that resulted in complete lysis of vaginal cells (overall score of 0+) but no lysis of sperm cells (overall score of 4+).

| DNA quantification
The slope, Y-intercept, and R 2 values were within the appropriate ranges and the amplification plots were of the expected curve morphology for all vaginal samples treated, regardless of method (data not shown).For groups with detectable DNA, mean DNA yields were compared across the cell lysis methods tested and no significant differences were observed (p = 0.818) (Figure 1).However, it should be noted that samples lysed with NaOH concentrations of ≥50 mM produced mean DNA yields (288.24ng-346.21ng) that were higher than the prepGEM™ control (197.24ng), while those lysed with 25 mM NaOH, high salt stain extraction buffer, urea/thiourea buffer, or digitonin exhibited mean DNA yields slightly lower F I G U R E 1 DNA yields obtained from vaginal eluates (n = 5) subjected to each cell lysis method.The prepGEM™ method was compared to all the other alternative cell lysis methods (n = 3 for urea/thiourea lysis buffer (30 min), n = 4 for digitonin lysis buffer, and n = 5 for alkaline buffer with 25-200 mM NaOH, high salt stain extraction buffer, mammalian protein extraction reagent (M-PER™), and urea/thiourea lysis buffer (15 min)).For modified RIPA lysis buffer (with proK), none of the five samples produced results, and thus this group was not included in the statistical evaluation.There was no significant difference observed between the lysis methods (p = 0.818).Boxes represent the upper and lower quartiles, spanning the interquartile range, with the median represented by the middle line.The whiskers represent the highest and lowest value obtained for each data set.
than the control (<150.49ng).Samples lysed with the M-PER™ reagent produced mean DNA yields equivalent to the control group, while those lysed with the modified RIPA lysis buffer (with proK) produced no detectable DNA (i.e., "undetermined").
In order to determine if the experimental lysis reagents were inhibiting the amplification reaction, IPCs were carefully evaluated across all methods tested (Table 1).The mean IPC C q values for all samples tested using alkaline (all NaOH concentrations tested) and M-PER™ reagents were within the acceptable range and were similar to the mean values obtained from the positive control group (p = 0.76).IPC amplification for samples lysed with high salt stain extraction buffer and digitonin lysis buffer crossed the threshold earlier than expected (resulting in lower C q values), indicating that the sample data is inconclusive.Finally, the other two lysis methods tested, urea/thiourea and modified RIPA (with proK), produced "undetermined" IPC values, which is consistent with inhibition.This is not surprising given that both contain known potentially inhibitory components; for example, the modified RIPA buffer contains sodium deoxycholate which is an ionic detergent that has been shown to inhibit PCR at high concentrations [43].Methods with this degree of inhibition would not be suitable for use in a forensic workflow.
The DI enables one to determine if the DNA sample exhibits any signs of degradation, which can be used to predict the quality of the downstream STR profile [44].The DI was calculated for each method to evaluate if any of the experimental reagents were potentially degradative to the DNA (Table 2).The DI for all samples treated with the control method (prepGEM™) were carefully examined across the individual donors to assure that each donor sample (as received) was of high quality.One of the donors had a DI of more than 10; therefore, it was removed from the data set (all methods) entirely and was not used for further testing.All other control samples showed no degradation, as expected.Vaginal samples lysed using all of the NaOH concentrations and M-PER™ displayed either no degradation or only mild degradation.Samples lysed using the urea/thiourea buffer (both time points) were highly degraded, which was not entirely unexpected given reports indicating that urea can cause oxidative stress, and thus DNA damage, at high concentrations [45].Alternately, samples lysed using the high salt stain extraction buffer and the digitonin lysis buffer produced low DIs along with low IPC C q values, indicating that the DI was uninterpretable.The samples lysed using the modified RIPA buffer (with proK) were so severely inhibited that DNA concentration values were not produced; consequently, these DIs were unable to be calculated (i.e., "undetermined").
Overall, the alkaline (25 mM, 50 mM, 75 mM, 100 mM, and 150 mM NaOH) and M-PER™ cell lysis methods generated high mean DNA yields, low DIs, and IPC C q values within the acceptable range; therefore, these were chosen for subsequent testing on semen samples.

| Preliminary evaluation of semen lysates
Prior to cell lysis, semen samples resulted in a mean score of 4+ showing numerous intact sperm cells.When used to lyse semen samples, the control lysis methods (prepGEM™ and 1 M NaOH) worked as expected; prepGEM™ -treated sperm cells were not lysed (score of 4+) and 1 M NaOH-treated sperm cells were fully lysed (score of 0) when examined microscopically (data not shown).For semen samples subjected to M-PER™ and alkaline lysis using lower concentrations of NaOH (e.g., 25 mM, 50 mM, 75 mM, and 100 mM), many intact sperm cells were observed across all donors (mean score of 4+), indicating little to no sperm cell lysis, as desired.However, alkaline lysis using 150 mM NaOH resulted in the visualization of a reduced number of sperm cells across the donors with a mean score of 2 + .TA B L E 1 IPC cycle threshold (C q ) values for vaginal lysates (n = 5) across the cell lysis methods.b n = 2.
Upon subsequent quantification, the DNA yields of experimental samples were compared to those obtained using the sperm lysis control method in order to determine the percentage of total DNA in each sample.Given that ~20% of DNA within normal semen originates from non-sperm cells [36][37][38], a viable method for lysing only non-sperm cells would show no more than ~20% of the total DNA in the semen sample lysate.Alkaline (25 mM NaOH) and M-PER™ lysis methods both resulted in DNA yields representing less than 20% of the total available DNA from the semen samples tested, which was consistent with the prepGEM™ control method (Figure 2).Alkaline lysis using higher concentrations of NaOH (e.g., 50 mM and 75 mM) resulted in lysis of a higher percentage of available cells (yielding 39.12 ± 18.12% and 40.45 ± 11.73% of total available DNA, respectively), indicating that these methods are capable of lysing some sperm cells.In addition, alkaline lysis using 100 mM and 150 mM NaOH resulted in significantly higher levels of sperm cell lysis when compared to the control method (49.19 ± 18.21% and 58.12 ± 9.28% vs. 9.57 ± 5.81% of total available DNA, respectively) (Figure 2, p < 0.001).Based on this quantitation data and subsequently produced STR profiles (data not shown), alkaline lysis methods using ≥50 mM NaOH were proven capable of lysing sperm cells, in addition to non-sperm cells, making them ineligible for differential cell lysis of forensic sexual assault samples.

| STR amplification of vaginal lysates
The primary goal of STR analysis was to identify non-sperm cell lysis methods that would produce 100% of expected STR alleles from vaginal samples.In addition, an ideal method would be one capable of producing STR profiles of equal or higher quality than the control lysis method (prepGEM™).All alternative lysis methods tested herein achieved 100% of expected STR alleles, except alkaline lysis using 25 mM NaOH, which produced 99 ± 1.15% of expected STR alleles (Table 3, p = 0.053).
F I G U R E 2 Percentage of total expected DNA obtained from semen using each of the non-sperm cell lysis methods.A significant difference was observed between the prepGEM™ control and alkaline lysis (both 100 mM NaOH and 150 mM NaOH) approaches (p < 0.001), demonstrating a greater degree of sperm cell lysis with those treatment conditions.No significant differences were observed between the control and all other experimental groups.The expected percentage of total DNA should be approximately 20% (denoted by the black line) since non-sperm cells only contribute 20% of the DNA in normal semen.
TA B L E 3 Summary of STR data for vaginal lysates (n = 4) from each cell lysis method.
In order to further examine the quality of the STR profiles produced from each experimental group, mean STR allele peak heights were calculated and compared to the control method (prepGEM™).
Although the mean STR allele peak heights from those samples processed using low concentration alkaline solutions (25 mM, 50 mM, and 75 mM NaOH) and the M-PER™ method were slightly lower than those processed with the control method, there were no significant differences between the groups (Table 3, p = 0.126).Moreover, intralocus balance (i.e., PHR) equal to or exceeding 0.8 was observed in all treatment groups, except for the 25 mM alkaline lysis technique which yielded a mean intralocus peak height ratio of 0.54 (Table 3).Finally, the interlocus balance was examined by calculating the CV of LPH:TPH for each sample and then averaging for each experimental group.
Based on the available literature for this method, a well-balanced profile would exhibit a CV of 0.35 or less [42].Peak heights across loci within a given profile should be similar in order to assure that all allele peaks are sufficiently above the analytical threshold so that true homozygosity can be confidently determined, as well as to enable more accurate differentiation of peaks from different individuals in a DNA mixture [46,47].Samples processed with the M-PER™ and all alkaline lysis methods (regardless of concentration) produced mean CVs that were not significantly different from that of the control group.
Overall, after assessing STR profiles obtained from vaginal samples processed using various alternative cell lysis techniques, none of the methods displayed a significant or consistent improvement over the control method (prepGEM™) in terms of all the metrics analyzed.However, lysates treated with the M-PER™ reagent exhibited DNA STR profiles of equivalent quality across all metrics examined (Table 3, Figure 3).Alkaline lysis (25 mM NaOH) also performed equivalent to the control across most metrics and should be considered a potentially viable method; however, further protocol optimization would be needed to address DNA quantification and STR quality issues noted.
In addition, there are other factors that must be considered for the implementation of a new direct-to-amplification non-sperm lysis technique, such as the time required to process samples and the cost per reaction (Table 4).Assuming a sample size of five, the entire traditional differential extraction process using the Qia-gen® QIAamp® DNA Investigator kit (Hilden; Germany) can take approximately three hours to complete and also relies heavily on manual pipetting and transfer steps.Furthermore, it is the most expensive reaction when compared to those used in this study.
One alternative to this method for non-sperm cell lysis is prep-GEM™; however, it is also a costly approach.On the other hand, the low concentration alkaline lysis method for non-sperm cell lysis used has an approximate cost of less than one U.S. dollar ($) per sample and the total processing time is about 15-20 min, and the M-PER™ method costs $0.17 per reaction and the total processing time is 35-40 min (Table 4).
Taken altogether, this study provides evidence that the M-PER™ method could be an alternative for non-sperm cell lysis, as this method consistently scored very well across all the quantitative and qualitative metrics examined herein and resulted in the best overall combination of tested performance and value.In addition, this new non-sperm cell lysis method, when combined with 1 M NaOH for sperm lysis, would potentially separate DNA fractions from sperm and non-sperm cells more efficiently than current methods.Ultimately, this would allow for the processing of more samples in less time at a reduced cost while potentially lowering the occurrence of DNA mixtures-increasing the certainty of STR profiling results and providing stronger scientific evidence in court.Further, given its ability to produce high-quality STR typing results rapidly without lysate purification, this method may offer a more viable option for transitioning the differential lysis approach onto a microdevice platform for even more time and cost savings.Going forward, testing with additional donors and replicates, including mock sexual assault mixture samples, would strengthen the implications of these findings.
Six different concentrations of NaOH reconstituted in MBG H 2 O (25 mM, 50 mM, 75 mM, 100 mM, 150 mM, and 200 mM) were tested on vaginal eluates, and five different concentrations of NaOH (25 mM, 50 mM, 75 mM, 100 mM, and 150 mM) were tested on semen samples.Ten microliters of the vaginal swab eluate or of the 1:10 semen was added to 11 μL of the specified NaOH concentration, and samples were incubated at 95°C for five minutes in 0.2 mL PCR tube strips using the ProFlex™ 3 × 32-well PCR System.Following incubation, 4 μL of 1 M Tris-HCl was added to the samples.

F I G U R E 3
Blue channels of representative electropherograms from vaginal eluates.Samples were subjected to: prepGEM™ (A), alkaline lysis using 25 mM NaOH (B), alkaline lysis using 50 mM NaOH (C), alkaline lysis using 75 mM NaOH (D), and M-PER™ (E).Data from other color channels were equivalent to that of the blue channel shown.
Degradation index observed for vaginal lysates.
Note:The value after ± denotes the standard deviation.TA B L E 2a n = 5.