Age estimation of burnt human remains through DNA methylation analysis.

The identification of human fire victims is a challenging task in forensic medicine. The heat-induced alterations of biological tissues can make the conventional anthropological analyses difficult. Even if the DNA profile of the victim is achieved, it is possible that no match can be found in a forensic DNA database, thus hindering positive identification. In such cases, any information useful to nail down a possible identity should be collected, such as DNA methylation analysis which could provide useful investigative leads. In the present study, five age-related epigenetic markers (ELOVL2, FHL2, KLF14, C1orf132, and TRIM59) were initially analysed in blood samples of 72 living Italian individuals of known age, using a Single Base Extension (SBE) assay. An age prediction model was built by multiple linear regression including all the markers (Mean Absolute Error, MAE: 3.15 years). This model was tested on 29 blood samples collected during autopsies from burnt human remains, already identified through DNA analysis, providing a MAE of 6.92 years. The model allowed a correct prediction in 79.3% of the cases (95% prediction interval), while six cases were associated with inaccurate predictions (min-max prediction error: 9.8-37.3 years). Among the different sample variables considered to explain these results, only the DNA degradation index was a relevant factor affecting the reliability of the predictions. In conclusion, the SBE typing of blood from burnt remains proved to be a reliable tool to estimate chronological age of most of the samples, also in consideration of its cost-effectiveness and the availability of CE sequencers in every forensic genetics laboratory.

Performance of the ForenSeqTM DNA Signature Prep kit on highly degraded samples.

Next generation sequencing (NGS) is the emerging technology in forensic genomics laboratories. It offers higher resolution to address most problems of human identification, greater efficiency and potential ability to interrogate very challenging forensic casework samples. In this study, a trial set of DNA samples was artificially degraded by progressive aqueous hydrolysis, and analyzed together with the corresponding unmodified DNA sample and control sample 2800 M, to test the performance and reliability of the ForenSeqTM DNA Signature Prep kit using the MiSeq Sequencer (Illumina). The results of replicate tests performed on the unmodified sample (1.0 ng) and on scalar dilutions (1.0, 0.5 and 0.1 ng) of the reference sample 2800 M showed the robustness and the reliability of the NGS approach even from sub-optimal amounts of high quality DNA. The degraded samples showed a very limited number of reads/sample, from 2.9-10.2 folds lower than the ones reported for the less concentrated 2800 M DNA dilution (0.1 ng). In addition, it was impossible to assign up to 78.2% of the genotypes in the degraded samples as the software identified the corresponding loci as "low coverage" (< 50x). Amplification artifacts such as allelic imbalances, allele drop outs and a single allele drop in were also scored in the degraded samples. However, the ForenSeqTM DNA Sequencing kit, on the Illumina MiSeq, was able to generate data which led to the correct typing of 5.1-44.8% and 10.9-58.7% of 58 of the STRs and 92 SNPs, respectively. In all trial samples, the SNP markers showed higher chances to be typed correctly compared to the STRs. This NGS approach showed very promising results in terms of ability to recover genetic information from heavily degraded DNA samples for which the conventional PCR/CE approach gave no results. The frequency of genetic mistyping was very low, reaching the value of 1.4% for only one of the degraded samples. However, these results suggest that further validation studies and a definition of interpretation criteria for NGS data are needed before implementation of this technique in forensic genetics.

Improving complex kinship analyses with additional STR loci.

In a standard paternity testing, mother, child, and alleged father are analyzed with STR markers using commercially available kits. Since Italian civil legislation does not have thresholds to confirm a paternity, paternity is practically proven when likelihood ratio increases prior probability of paternity to posterior, accepted by court as sufficient. However, in some cases the number of markers included in a commercial kit may be insufficient to conclusively prove or disprove a relationship between individuals, especially when complex family scenarios are suspected or indirect analyses are required. Additional genetic information can increase the values of the likelihood ratio regarding the detection of true parental relationships in a pedigree, while reducing the chances of false attributions (e.g. false paternities). In these cases the introduction of a 26Plex amplification system allows to examine 23-26 additional markers depending on the commercial kit used, thus increasing the statistical power of the kinship analysis. The PCR conditions were optimized for a multiplex amplification system and a new generation CE instrument. In order to demonstrate the utility of additional STRs markers, four complex kinship cases are presented.

The molecular characterization of a depurinated trial DNA sample can be a model to understand the reliability of the results in forensic genetics.

The role of DNA damage in PCR processivity/fidelity is a relevant topic in molecular investigation of aged/forensic samples. In order to reproduce one of the most common lesions occurring in postmortem tissues, a new protocol based on aqueous hydrolysis of the DNA was developed in vitro. Twenty-five forensic laboratories were then provided with 3.0 µg of a trial sample (TS) exhibiting, in mean, the loss of 1 base of 20, and a molecular weight below 300 bp. Each participating laboratory could freely choose any combination of methods, leading to the quantification and to the definition of the STR profile of the TS, through the documentation of each step of the analytical approaches selected. The results of the TS quantification by qPCR showed significant differences in the amount of DNA recorded by the participating laboratories using different commercial kits. These data show that only DNA quantification "relative" to the used kit (probe) is possible, being the "absolute" amount of DNA inversely related to the length of the target region (r(2) = 0.891). In addition, our results indicate that the absence of a shared stable and certified reference quantitative standard is also likely involved. STR profiling was carried out selecting five different commercial kits and amplifying the TS for a total number of 212 multiplex PCRs, thus representing an interesting overview of the different analytical protocols used by the participating laboratories. Nine laboratories decided to characterize the TS using a single kit, with a number of amplifications varying from 2 to 12, obtaining only partial STR profiles. Most of the participants determined partial or full profiles using a combination of two or more kits, and a number of amplifications varying from 2 to 27. The performance of each laboratory was described in terms of number of correctly characterized loci, dropped-out markers, unreliable genotypes, and incorrect results. The incidence of unreliable and incorrect genotypes was found to be higher for participants carrying out a limited number of amplifications, insufficient to define the correct genotypes from damaged DNA samples such as the TS. Finally, from a dataset containing about 4500 amplicons, the frequency of PCR artifacts (allele dropout, allele drop-in, and allelic imbalance) was calculated for each kit showing that the new chemistry of the kits is not able to overcome the concern of template-related factors. The results of this collaborative exercise emphasize the advantages of using a standardized degraded DNA sample in the definition of which analytical parameters are critical for the outcome of the STR profiles.

Looking into the Quantification of Forensic Samples with Real-Time PCR.

The quantification of human DNA extracts from forensic samples plays a key role in the forensic genetics process, ensuring maximum efficiency and avoiding repeated analyses, over-amplified samples, or unnecessary examinations. In our laboratory, we use the Quantifiler® Trio system to quantify DNA extracts from a wide range of samples extracted from traces (bloodstains, saliva, semen, tissues, etc.), including swabs from touched objects, which are very numerous in the forensic context. This method has been extensively used continuously for nine years, following an initial validation process, and is part of the ISO/IEC 17025 accredited method. In routine practice, based on the quantitative values determined from the extracts of each trace, we use a standard method or a low-copy-number method that involves repeating the amplification with the generation of a consensus genetic profile. Nowadays, when the quantification results are less than 0.003 ng/µL in the minimum extraction volume (40 µL), we do not proceed with the DNA extract analysis. By verifying the limits of the method, we make a conscious cost-benefit choice, in particular by using the least amount of DNA needed to obtain sufficiently robust genetic profiles appropriate for submission to the Italian DNA Forensic Database. In this work, we present a critical re-evaluation of this phase of the method, which is based on the use of standard curves obtained from the average values of the control DNA analysed in duplicate. Considering the various contributions to uncertainty that are difficult to measure, such as manual pipetting or analytical phases carried out by different operators, we have decided to thoroughly investigate the contribution of variability in the preparation of calibration curves to the final results. Thus, 757 samples from 20 independent experiments were re-evaluated using two different standards for the construction of curves, determining the quantitative differences between the two methods. The experiments also determined the parameters of the slope, Y-intercept, R2, and the values of the synthetic control probe to verify how these parameters can provide information on the final outcome of each analysis. The outcome of this revalidation demonstrated that it is preferable to use quantification ranges rather than exact quantitative limits before deciding how to analyse the extracts via PCR or forgoing the determination of profiles. Additionally, we present some preliminary data related to the analysis of samples that would not have been analysed based on the initial validation, from which genetic profiles were obtained after applying a concentration method to the extracts. Our goal is to improve the accredited analytical method, with a careful risk assessment as indicated by accreditation standards, ensuring that no source of evidence is lost in the reconstruction of a criminal event.

Genetic analysis of suicide: a sample study in Tuscany (Central Italy).

Many studies have examined the genetic contribution to suicide. However, data on suicide in the Italian population are scarce. We therefore aimed to address this gap by investigating a cohort of 111 Italians for whom a verdict of suicide had been declared in court in Florence, Italy between 2007 and 2017. This cohort included 86 men and 25 women. DNA samples were obtained from tissues or blood, and 22 genes from multiple neurobiological pathways previously shown to be associated with the pathogenesis of suicide were analysed. Next-generation sequencing was used to compare these gene sequences with those from a large, normal population. In this study, we identified 19 gene variants that were present at significantly lower frequencies in our Italian cohort than in the general population. In addition, four missense mutations were identified in four different genes: Monoamine Oxidase A (MAOA), 5-Hydroxytryptamine Receptor 2 A (HTR2A), Sodium Voltage-Gated Channel Alpha Subunit 8 (SCN8A), and Nitric Oxide Synthase 3 (NOS3). Our study identified several potential genetic links with suicide in a cohort of Italians and supports a relationship between specific genetic variants and suicidal behaviour in this population.

Estimating the integrity of aged DNA samples by CE.

A CE/UV method was developed to separate by a micellar system the four DNA bases and other five purinic-pyrimidinic compounds (5-methyl-cytosine, uracil, xanthyne, hypoxanthyne and 5-bromo-uracil). Selectivity, precision, accuracy and sensitivity were assessed and proved to be suitable for the analysis of the primary structure of DNA. This method was adopted to study 16 aged samples including two Egyptian mummies, formaldehyde-fixed paraffin-embedded tissues and other forensic specimens. Lower relative values of the four canonical unmodified DNA bases (uDNAb) and more complex pherograms were found in the aged samples when compared with the modern controls. The results of the CE analysis, together with those obtained by classical molecular methods (agarose gel electrophoresis, DNase I and RNase A assays, and UV spectrophotometry), were finally evaluated for assessing the reliability of STR typing. Since samples with low uDNAb showed no amplification or unreliable STR profiles, the uDNAb value is discussed as a further quality criterion in the evaluation of the genetic data obtained from aged samples.

A single mutation in the FGA locus responsible for false homozygosities and discrepancies between commercial kits in an unusual paternity test case.

We report an unusual paternity test case showing multiple peculiarities. Using AmpFlSTR Profiler Plus and AmpFlSTR Identifiler PCR Amplification kits, the alleged father and the two children were apparently homozygous at the FGA locus, but using the PowerPlex 16 kit the three individuals were found to be heterozygous. Drop-out was caused by a single mutation event in the presumptive binding site of the reverse primer. In addition, three inconsistencies were detected between the daughter and the alleged father among 18 STR markers. The occurrence of the rare null allele at the FGA locus and case history suggested that the true father was the brother of the alleged father. Furthermore, a single-step repeat maternal mutation was also detected at D16S539. This puzzling case was solved by using multiple analytical approaches, including the use of different primer pairs, the use of a high number of STR markers, and the characterization of the mutation causing the "null allele."

Validation of a large Italian Database of 15 STR loci.

Results from a collaborative exercise with proficiency testing conducted by 20 Italian laboratories on the 15 loci included in the Identifiler kit were analyzed by allele sharing methods and by standard population genetics tests. The validated database, including about 1500 subjects, was merged with that of a previous exercise conducted on nine loci, and the resulting allele frequencies, subdivided by Italian region, were published on-line.

Analysis of 13 tetrameric short tandem repeat loci in a population of Tuscany (Central Italy) performed by means of an automated infrared sequencer.

Allele frequencies for the 13 STRs of the Combined DNA Index System (CODIS) core were obtained from a sample of 188 unrelated individuals living in the area of Florence, Prato and Pistoia (Tuscany, Central Italy).

Allele sharing in first-degree and unrelated pairs of individuals in the Ge F I AmpFlSTR Profiler Plus database.

Eleven Italian forensic laboratories participated in a population study based on the AB Profiler Plus loci with proficiency testing. The validated database, including 1340 individuals, is available on-line. Tests for Hardy-Weinberg equilibrium, gametic unbalance, and heterogeneity of gene frequency were generally not significant. Gene frequencies at each locus were consistent with those of two previously published Italian studies, but different from a third. Individuals of each subsample were paired, and the total number of alleles shared across the nine loci was determined in each pair. The analysis was replicated over the total sample. In addition, two samples of mother-child pairs (N=315) and full-sib pairs (N=91) were subjected to allele sharing analysis. The resulting distributions were sufficiently distinct from the sample of unrelated pairs as to be of practical usefulness.

Stability of human α-salivary amylase in aged forensic samples.

The unequivocal tissue identification in forensic casework samples is a key step for crime scene reconstruction. Just knowing the origin of a fluid can sometimes be enough to either prove or disprove a fact in court. Despite the importance of this test, very few data are available in literature concerning human saliva identification in old forensic caseworks. In this work the stability of human α-amylase activity in aged samples is described by using three different methods integrated with DNA profiling techniques. This analytical protocol was successfully applied on 26-years old samples coming from anonymous threat letters sent to prosecutors who were working on "the Monster of Florence", a case of serial murders happened around Florence (Italy) between 1968 and 1985.

False-positive results with amylase testing of citrus fruits.

In a case of robbery in which the criminals passed through the garden adorned with calamondin trees (Citrus madurensis), the investigators found in the grass six calamondin fruits, some undamaged, while others apparently bitten. The fruits were collected and sent to the laboratory for DNA analysis to verify the presence of saliva and robbers' DNA profile. A specific immunochromatographic strip test for saliva confirmed the presence of human salivary α-amylase, but similar positive results were also observed for intact calamondin and other citrus fruits. Further analysis with a specific automated amylase test confirmed the absence of amylase activity. DNA quantification and typing using a specific forensic kit revealed no human DNA presence in any fruits. This case report demonstrates for the first time the occurrence of false positives when human saliva is sought on citrus fruits.

Genetic profiling of Bolivian population using 15 STR markers of forensic importance.

Allele frequencies for the 13 CODIS markers plus D2S1358 and D19S433 loci were estimated in a sample of 148 unrelated individuals from Bolivia and parameters of forensic interest were calculated. Further, the STR data were analyzed using a distance-based method to assess the genetic relationships of this population with other ones living in Argentina, Brazil, Costa Rica, Mexico, Peru, Venezuela and three autochthonous populations living in the Beni Department of Bolivia (Quechua, Aymara and Beni population).

Type A microsatellite instability in pediatric gliomas as an indicator of Turcot syndrome.

Microsatellite instability (MSI) is present in hereditary conditions due to mismatch repair (MMR) gene mutations. Following MSI analysis, tumor samples are classified into MSS (stable), MSI-L (low instability), and MSI-H (high instability) based on the fraction of unstable loci. Another MSI-based classification takes into account the size difference between mutant alleles in tumor DNA compared to wild-type alleles; two types of MSI, A and B, are recognized using this approach, type A being characterized by smaller, more subtle allelic shifts compared to type B. Biallelic mutations of MMR genes are associated with pediatric cancers, including glial tumors, in Turcot syndrome type 1 (TS1). However, most TS1-associated gliomas so far analyzed did not display MSI. We investigated the frequency of MSI in a series of 34 pediatric gliomas of different grade using a panel of five mononucleotide quasimonomorphic markers. Subtle qualitative changes were observed for the majority of markers in two glioblastomas (5.9% of the total series and 33.3% of glioblastomas). In both cases, family histories were compatible with TS1, and mutations of the PMS2 and MLH1 genes were identified. In one family, the MSI patterns were compared between the glioblastoma and a colon cancer from an affected relative, showing a clear qualitative difference, with the former displaying type A and the latter type B instability, respectively. These results were confirmed using additional microsatellite markers, indicating that knowledge of the association between TS1-related glial tumors and subtle type A MSI is important for full ascertainment of TS1 patients and appropriate counselling.

Genetic STRs variation in a large population from Tuscany (Italy).

Allele frequencies for 17 STRs, together with some parameters of forensic interest, were estimated in a sample of 835 unrelated individuals born in Tuscany, an Italian region. These data were compared with Italian, Chinese, Kosovo Albanian, Romanian and Tunisian populations, strongly represented in this area. No significant differences in single loci were detected, except for Chinese in comparison with all the other populations.

Mixed stains from sexual assault cases: autosomal or Y-chromosome short tandem repeats?

We analyzed forensic DNA samples from four cases of sexual assault, using the Y-chromosome-specific human DNA markers and a panel of autosomal short tandem repeats (STRs). The presence of male contribution was evaluated by the analysis of the Amelogenin locus. A panel of tetrameric Y-STR (DYS19, DYS390, DYS391, DYS392, DYS385, DYS389I and II) was used in further analysis of samples, increasing the efficiency of the forensic genetic analyses. It was possible to identify a partial or full Y-profile of the rapists in different DNA mixtures when genetic profile could not be detected by autosomal STRs. However, in the case of male/male DNA mixture, only the victim's Y-profile could be obtained because the DNA of the offenders was present in low amounts. When the mixture contained different male/male proportion of DNA, only the full profile of the major component could be detected. In cases where male/female DNA mixed stains contained a sufficient amount of male DNA, the analysis of autosomal STRs was adequate enough to identify the full profile of the rapist. Our experience shows that the main advantage of the Y-STR approach is its ability to detect the male component in the mixed stains when the DNA of the male contributor is present only in a very small amount.

Double incompatibility at human alpha fibrinogen and penta E loci in paternity testing.

We present a paternity testing case in which a double incompatibility was found for two short tandem repeat (STR) markers, human fibrinogen alpha (FGA) and Penta E. Analysis of the trio (mother, father, and daughter) included the amplification with a battery of 15 autosomal short tandem repeats (STR) by using a commercially available PowerPlex 16 System kit, and the detection with an ultraviolet-automatic sequencer. The biological paternity was confirmed with 12 additional markers. Reanalysis of the trio for the same markers with different primers was carried out by using an infrared automated sequencer and infrared-fluorescent primers. High paternity index confirmed that the observed inconsistencies were due to a double mutation, which was confirmed by sequence analysis at FGA and Penta E loci. Amplification and detection results obtained by the infrared-protocol showed consistent results with those obtained by ultraviolet-protocol and a commercially available kit. This has been our first case of double mutation at FGA and Penta E in a paternity testing. The use of our approach, based on two amplification and detection formats and on the sequence analysis, confirmed the observed meiotic paternal mutations.

Haplotype studies support slippage as the mechanism of germline mutations in short tandem repeats.

Germline mutations of human short tandem repeat (STR) loci are expansions or contractions of repeat arrays which are not well understood in terms of the mechanism(s) underlying such mutations. Although polymerase slippage is generally accepted as a mechanism capable to explain most features of such mutations, it is still possible that unequal crossing over plays some role in those events, as most studies in humans could not exclude unequal crossing over (UCO). Crossing over can be studied by analyzing haplotypes using flanking markers. To check for UCO in mutations, we have analyzed 150 paternity cases for which more than the usual trio (mother, child, and father) were available for testing by analyzing 16 STR loci. In a total of 4900 parent-child allele transfers four mutations were observed at different loci (D8S1179, D18S51, D21S11, and SE33/ACTBP2). To identify the mutated allele and to check for UCO, we typed at least four informative loci flanking the mutated locus and used the pedigree data to establish haplotypes. By doing so we were able to exclude UCO in each case. Moreover, we were able to identify the mutations as one-repeat contractions/expansions. Our data thus support slippage as the mechanism of germline mutations in STRs.

Identification of forensic samples by using an infrared-based automatic DNA sequencer.

We have recently introduced a new protocol for analyzing all core loci of the Federal Bureau of Investigation's (FBI) Combined DNA Index System (CODIS) with an infrared (IR) automatic DNA sequencer (LI-COR 4200). The amplicons were labeled with forward oligonucleotide primers, covalently linked to a new infrared fluorescent molecule (IRDye 800). The alleles were displayed as familiar autoradiogram-like images with real-time detection. This protocol was employed for paternity testing, population studies, and identification of degraded forensic samples. We extensively analyzed some simulated forensic samples and mixed stains (blood, semen, saliva, bones, and fixed archival embedded tissues), comparing the results with donor samples. Sensitivity studies were also performed for the four multiplex systems. Our results show the efficiency, reliability, and accuracy of the IR system for the analysis of forensic samples. We also compared the efficiency of the multiplex protocol with ultraviolet (UV) technology. Paternity tests, undegraded DNA samples, and real forensic samples were analyzed with this approach based on IR technology and with UV-based automatic sequencers in combination with commercially-available kits. The comparability of the results with the widespread UV methods suggests that it is possible to exchange data between laboratories using the same core group of markers but different primer sets and detection methods.