Short tandem repeat profiling via next-generation sequencing for cell line authentication

Abstract

Cell lines are indispensable models for modern biomedical research. A large part of their usefulness derives from the ability of a cell line to proliferate over multiple passages (often indefinitely), allowing multiple experiments to be performed. However, over time, cell line identity and purity can be compromised by human errors. Cross-contamination from other cell lines and complete misidentification are both possible. Routine cell line authentication is a necessary preventive measure and has become a requirement for many funding applications and publications. Short tandem repeat (STR) profiling is the most common method for cell line authentication and is usually carried out using standard polymerase chain reaction-capillary electrophoresis analysis (STR-CE). Here, we evaluated next-generation sequencing (NGS)-based STR profiling of human and mouse cell lines at 18 and 15 loci, respectively, in a high-throughput format. Using the Python program STRight, we demonstrate that NGS-based analysis (STR-NGS) is superior to standard STR-CE in terms of the ability to report the sequence context of repeat motifs, sensitivity and flexible multiplexing capability. STR-NGS is thus a valuable alternative for cell line authentication.

Keywords: Capillary electrophoresis; Cell identity; Cell line authentication; Next-generation sequencing; Short tandem repeat; Targeted deep sequencing.

Fig. 1.

STR-NGS optimization and performance. (A) Schematic representation of a short tandem repeat (STR) locus with STR repeats, flanking regions at the start and end of the repeat region, and a targeted deep sequencing (DS) primer pair. Sequence from ‘5′ Flank’ to ‘3′ Flank’ is used for repeat length analysis. (B) Allele and stutter frequencies for five representative STR loci calculated as the percentage of the parent allele reads. Verified STRs are indicated by a ‘#’ above the bar plot. Standard and optimized conditions are MyTaq and SuperFi+tetramethylammonium oxalate (TMAO), respectively. (C) Total background (stutter+noise) in all STR loci in the comparison between standard and optimized PCR conditions. (D) STR profile for each locus examined by PCR-capillary electrophoresis (CE)- and next-generation sequencing (NGS)-based methods in two diploid induced pluripotent stem cell (iPSC) lines. Repeat lengths containing a decimal point indicate that an additional partial repeat is present in the locus. Red text indicates STRs that differed between STR-CE and STR-NGS. N/A, not available with STR-CE.

Fig. 2.

STR-NGS sensitivity on a mixed sample and optimized multiplexed conditions. (A) Observed allele fractions of informative STRs repeats are plotted against the expected ratio for given mixtures of two genomic samples. AN1.1 gDNA was diluted into BJFF.6 gDNA in a ratio of 1:1 to 1:1000. Expected allele fractions for diluted AN1.1 STRs correlate well with observed allele fractions. (B) A mixture of two diploid cell lines was analyzed for STR D8S1179. ‘M’ and ‘N’ alleles indicate genotypes from the major and minor components, respectively. Bar graphs shows percentage read counts for each STR repeat of both cell lines in different mixture ratios. (C) STR profile for each locus examined using multiplexed STR-NGS in BJFF.6 iPSC cells. The asterisk indicates a dropout allele.

Fig. 3.

STR-NGS in mouse cell lines. (A) STR repeat length profiles in two mouse cell lines (NIH3T3 and MC3T3-E1) show allele, stutter and noise fractions of total parent allele reads. The asterisk indicates the difference in STR repeat length between reference and STR-NGS. (B,C) The frequencies of different stutters and repeat lengths (B), and sequence context of different STR repeat lengths (C) in two STR loci (4-2 and 18-3) are different between reference and STR-NGS in NIH3T3 cells. Repeat lengths containing a decimal point indicate that a partial repeat is present in predominant read length.