Introduction

DNA analysis with short tandem repeat (STR) loci presently constitutes the main human identification (HID) tool all around the world [1]. However, replication and disjunction inaccuracies along with technical pitfalls to analyze the human genome can modify the well-known Mendelian inheritance pattern, which complicates the interpretation to establish kinship biological relationships [2]. For instance, the relative high mutation rate of the STR loci is frequently responsible of paternity test exclusions. However, knowledge of biological processes causing STR mutations, such as the replication slippage framed into the stepwise mutation model (SMM) [3], allows recognizing and interpreting these findings [4, 5]. Among the technical pitfalls during the HID process, mutations in the primer annealing site can prohibit the PCR amplification of some alleles, also known as null alleles [2]. An alleged father carrying null alleles in HID markers will be detected as a false homozygous, and he could be excluded in a paternity test when he passes on the null allele to his child, who will appear as homozygous for the maternal allele. Similar to the incompatibilities by mutation, statistical approaches have been implemented to interpret the null allele presence in paternity tests [6]. However, some atypical situations arise during the continuous resolution of paternity cases, which constitute challenges that require the inclusion of additional genetic systems and non-traditional interpretation. In this work, we report one paternity case where a deeper genetic and statistical analysis was necessary to explain the paternal relationship despite the presence of three father-child incompatibilities, which allowed suggesting a maternal uniparental isodisomy for the chromosome 2.

Material and methods

DNA was extracted using the DNA IQ system kit (Promega Corp). Genotypes were obtained for a total of 24 STR loci included in the Powerplex® Fusion and/or Globalfiler systems, namely D3S1358, D1S1656, D2S441, D10S1248, D13S317, PENTA E, D16S539, D18S51, D2S1338, CSF1PO, PENTA D, TH01, vWA, D21S11, D7S820, D5S818, TPOX, DYS391, D8S1179, D12S391, D19S433, FGA, D22S1045, and SE33, plus amelogenin. Due to the presence of two incompatibilities (TPOX, D2S441), the PowerplexY-23 system was analyzed in the father and child including the following Y-STRs: DYS19, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, DYS385a/b, DYS439, DYS437, DYS456, DYS635, DYS438, GATA-H4, DYS448, DYS481, DYS549, DYS553, DYS643, DYS458, DYS570, and DYS576. We followed the amplification and capillary electrophoresis conditions recommended by the suppliers, using the corresponding allelic ladders. For the genotyping process, the softwares GeneMapper v3.2 (Applied Biosystems, Foster City, CA) and GeneMapper® ID-X v1.4 (Thermo Fisher Scientific) were used in the ABI Prism 3130 and 3500 Genetic Analyzers, respectively. For deeper analysis, the genotype for 38 autosomal HID indels was obtained according to the protocol described by Pereira et al. (2009) [7]. Alleged father and mother signed a written informed consent authorizing the publication of this case; personal data will be preserved at all time.

Paternity indexes (PI) were computed with the Familias 3.0 software [8]. For this purpose, we used allele frequencies of population studies in Mexico reporting forensic parameters for both Powerplex® Fusion and Globalfiler kits [9], and for the 38 HID-indel system [10], respectively. In the software Familias 3.0, the mutation probability was evaluated only in two STR loci (TPOX and D2S441) by means of the SMM (unstationary) with default parameters (range 0.1; rate 21e−006) and an overall mutation rate of 0.005. Similarly, the null allele probability used the minimum allele frequency (5/2N) for estimations. The paternity indexes (PI) were estimated from the comparison of probabilities for three hypothesis: (1) Alleged father (AF) is the father of the child, (2) a random man is the father of the child, (3) A brother of the AF is the father of the child. Equal prior probabilities (0.5) for and against paternity were assumed to calculate the respective probabilities of paternity (W). Finally, we used the YHRD tool available in this database to evaluate the kinship index based on the Y-STR haplotype (https://yhrd.org/kinship) [11].

Results

A DNA paternity test was carried out with the autosomal STR kits, which displayed two father-child incompatibilities, for TPOX (11/11 vs 8/8) and D2S441 (14/14 vs 10/10) (Table 1). However, following the International Society of Forensic Genetics (ISFG) recommendations [12], we estimated the residual PI indicating a real kinship relationship assuming the probable presence of mutations/null alleles (PI = 2.2107e+8) (Table 1). Although the presence of few incompatibilities also could be explained because the brother of the AF is the father of the child, this hypothesis was lesser probable (PI = 2.8358e+7). The biological relationship between the AF and child also was paternally corroborated with the PowerplexY-23 system (Supplementary Table S1). Although the father and child homozygosity suggests the null allele presence, amplification with different autosomal HID kits did not support this possibility. In addition, these null alleles are not listed in the STRbase of the National Institute of Standard and Technologies (http://strbase.nist.gov/NullAlleles.htm). Similarly, although multistep STR mutations are underrepresented due to the classification bias toward shorter mutations [13], two simultaneous incompatibilities would be poorly explained by mutation because they would imply three and four mutational steps, respectively [2]. Conversely, the two observed incompatibilities and the homozygous state of the child for three STRs located in the chromosome 2 (TPOX, D2S441, and D2S1338) suggest that uniparental isodisomy is the best explanation for this case. It must be noticed that the child shares one allele with his mother for every locus (Table 1). In order to confirm this hypothesis, we obtained the genotype for the 38 HID indel systems (Supplementary Table S2). Interestingly, the alleged father-child incompatibilities displayed one exclusion in the unique locus of this genetic system located in the chromosome 2 (B02 2/2 vs. 1/1, respectively). Once more, the PI of the AF was larger than those of the AF brother hypothesis (PI = 874.6 vs 112.9, respectively), and the child was homozygous for one maternal allele, supporting the maternal uniparental isodisomy hypothesis for this case.

Table 1 Genotype for 23 autosomal STRs obtained with the Powerplex Fusion and Globalfiler kits of the pedigree with probable maternal uniparental disomy in the chromosome 2
Full size table

Based on all the 61 autosomal loci analyzed herein, we contrasted the random man versus AF and AF brother hypotheses, which supported—again—that AF is the father of the child (PI = 1.94e+11 vs. 3.20e+09, respectively). In posterior probability (W), these values represent 98.4% for the AF paternity, regarding 1.6% for the AF brother paternity (Table 2). Considering that AF is the father, because brothers share the same Y-haplotype, the AF paternity probability would be almost 100% when the Y-linked genetic evidence is added to the autosomal evidence (PI = 6.0795e+15) (Table 2). Combination of information from autosomal and Y-chromosome markers is in agreement with the ISFG recommendations [12]. Conversely, the probability to observe three null alleles was estimated in 3.099e−8 based on minimum allele frequencies reported in population studies [9, 10], and that each null allele was passed on from the father to the child (0.53). This negligible probability indicates that the maternal uniparental isodisomy hypothesis is more plausible in this case. Interestingly, the entire chromosome 2 probably is implied given that the incompatible loci are in p and q arms (Fig. 1).

Table 2 Global paternity index obtained considering the null allele presence
Full size table
Fig. 1
figure 1

Chromosomal position of the loci analyzed in this paternity test where child is homozygous for maternal alleles, suggesting the entire chromosome 2 is involved in the maternal uniparental isodisomy (MUPD)

Full size image

Uniparental isodisomy cases in paternity testing have been scarcely reported in the literature, particularly for smaller chromosomes, such as the 6 [14], 16 [15], and 21 [16]. To our knowledge, for the chromosome 2, only one partial maternal uniparental isodisomy [17] and one entire paternal uniparental isodisomy [18] cases have been reported, without apparent clinical problems [19]. Unfortunately, although clinical evaluation would be recommended in this case, the family was not available for additional follow-up. In brief, the genetic evidence obtained in this paternity case confirmed a father-child relationship despite the presence of three autosomal incompatibilities. Deeper analysis of these incompatibilities involving markers located in the chromosome 2 allowed concluding a maternal uniparental isodisomy as the more plausible explanation for these results.