Oligonucleotide (CAC)5 fingerprinting: validity and reliability in paternity testing

Conformational properties of DNA containing (CCA)n and (TGG)n trinucleotide repeats

We have used CD spectroscopy, polyacrylamide gel electrophoresis, and UV absorption spectroscopy to study conformational properties of DNA fragments containing (CCA)n and (TGG)n repeats, which are the most length-polymorphic microsatellite sequences of the human genome. The (CCA)n fragments are random single strands at neutral and alkaline pH but they fold into intramolecular intercalated cytosine tetraplexes at mildly acid pH values. More acid values stabilize intermolecular tetraplex formation. The behavior of (TGG)n repeats is more complex. They form hairpins or antiparallel homoduplexes in low salt solutions which, however, are transformed into parallel-stranded guanine tetraplexes at physiological KCl concentrations. Their molecularity depends on the repeat number: (TGG)4 associates into an octameric complex, (TGG)8 forms tetramolecular complexes. (TGG)n with odd repeat numbers (5, 7, and 9) generate bimolecular and tetramolecular tetraplexes. The only (TGG)7 folds into an intramolecular tetraplex at low KCl concentrations, which is antiparallel-stranded. Moreover, the (TGG)(n) fragments provide various mutually slipped conformers whose population increases with salt concentration and with the increasing repeat number. However, the self-structures of both strands disappear in the presence of the complementary strand because both (TGG)n and (CCA)n prefer to associate into the classical heteroduplex. We suppose that the extreme conformational variability of the DNA strands stands behind the length polymorphism which the (CCA)n/(TGG)n repeats exhibit in the human genome.

Phenotypic and genotypic characterization of commonly used human prostatic cell lines

OBJECTIVE

To investigate and catalogue systematically the phenotypic and genotypic characteristics of the commonly used prostatic cell lines using immunocytochemistry and polymerase chain reaction (PCR) of hypervariable sequences within the genome to provide a 'fingerprint' characteristic of each cell line. Materials and methods Malignant (LNCaP, LNCaP-r, PC-3, DU-145) and benign immortalized prostatic cell lines (PNT-1A, PNT-1B, BPH-1) were grown on four-well slides, fixed and subjected to indirect streptavidin-biotin immunocytochemistry. Twenty-three antibodies were used in the following groups: cytoskeletal elements: cytokeratins (CK)-5, -7, -8, -14 (two), -16, -18, -19 (three), -20, vimentin and desmin; MUC1 (three); cell adhesion molecules (E-cadherin, alpha-beta-and gamma-catenin); and prostatic associated proteins: prostate specific antigen (PSA), prostatic acid phosphatase (PAP) and androgen receptor (AR). For the PCR, genomic DNA was extracted from the cell lines and from SKOV3 and MCF7 (positive controls). PCR was performed on three variable regions which were then sequenced: AR exon 1 (CAG repeat polymorphism), and two areas of microsatellite instability (MSI): AR exon 8 and hypoxanthine-guanine phosphoribosyl transferase (HPRT) exon 3.

RESULTS

All cell lines were CK-8/18 positive and most also expressed CK-7 and -19. Heterogeneous CK-20 expression was detected for the first time in prostatic cell lines. All lines were positive for vimentin and negative for desmin. MUC1 was expressed in one malignant (DU-145) and all immortalized cell lines. E-cadherin expression was low or absent in three lines: PNT1A, 1B and PC-3. Only PC-3 failed to express alpha-catenin; beta- and gamma-catenin were expressed by all lines. PSA, PAP and AR were only expressed by LNCaP and LNCaP-r. On PCR, the CAG repeat lengths in exon 1 of the AR ranged from 19 to 27. Three pairs of cell lines had the same exon 1 CAG repeat length: LNCaP/PC-3 (26 repeats), BPH-1/DU-145 (19 repeats) and PNT1 A/1B (20 repeats). Exon 8 sequences were identical except for LNCaP, which showed a single base mutation, and HPRT exon 3 sequences were all identical. There was no evidence of generalized MSI in any of the cell lines examined.

CONCLUSIONS

The cell lines studied fell into three broad groups according to their phenotypic characteristics: (i) prostatic marker positive (LNCaP and LNCaP-r); (ii) high expression of most antigens (DU-145, PC-3 and BPH-1); and (iii) low or absent expression of most antigens (PNT1 A and 1B). Each of the cell lines derived from PC could be identified on the basis of exon 1 and 8 AR sequence variability. DU145 and BPH-1 had identical profiles of the three areas studied, but these cell lines are easily distinguished by their different phenotypic characteristics. PNT1A and 1B had identical genetic and similar phenotypic profiles, which is unsurprising given that they are subclones derived from the same parental line. Even so, these were separable on the basis of CK-19 immunostaining. Using a combination of geno- and phenotypic markers it was possible to derive a 'fingerprint' for each of the cell lines assessed, which will allow meaningful comparison between similar cell lines held in other laboratories.