Pancreatic Cancer Cell Lines Heterozygous for the SPINK1 p.N34S Haplotype Exhibit Diminished Expression of the Variant Allele
The c.101A>G (p.N34S) variant and its associated haplotype in the serine protease inhibitor Kazal-type 1 (SPINK1) gene is one of the clinically most significant risk factors for chronic pancreatitis.1,2 Although a loss-of-function mechanism has been proposed for the increased pancreatitis risk conferred by this haplotype, experimental evidence is lacking because neither the p.N34S variant nor any of the 4 associated intronic variants affect SPINK1 expression or trypsin inhibitory function.3–8 Recently, 2 pancreatic adenocarcinoma cell lines, PaCa44 and PancTu-I, were reported to carry a heterozygous p.N34S variant.9 The aim of the present study was to take advantage of the unique opportunity that these cell lines offer and investigate whether the p.N34S haplotype affects messenger RNA (mRNA) expression of SPINK1.
PaCa44 and PancTu-I cells10 were kind gifts from Matthias Löhr (Karolinska Institutet) and Ralf Jesenofsky (University of Heidelberg). Both cell lines expressed SPINK1 mRNA detectable by reverse transcription polymerase chain reaction (PCR). Heterozygosity for the p.N34S variant was confirmed by sequencing both genomic DNA and complementary DNA (cDNA) (Fig. 1A). Relative expression levels of the 2 SPINK1 alleles were determined using reverse transcription PCR followed by allele-specific digestion with the restriction endonuclease Hpy166II and quantitation of the digestion products (Figs. 1B, C). Remarkably, in this experiment, the p.N34S allele showed reduced expression, which was 6.7-fold lower than the expression of the wild-type allele (Fig. 1D). Although not shown, additional experiments using different primer sets showed 3-fold to 4-fold reduced expression of the variant allele. The observations suggest that the p.N34S haplotype may comprise a negative regulatory variant likely located in the 5′ region upstream of the gene. To search for potential candidates, we determined the genomic DNA sequence for the entire SPINK1 gene (6.9 kb) and the flanking 5′ (6.2 kb) and 3′ (1.2 kb) regions. We identified 22 variations relative to the reference sequence, which were, unexpectedly, identical in the 2 cell lines (Table 1). The 5′ region contained 6 single-nucleotide variants. When subjects homozygous (n = 3) or heterozygous (n = 3) for the p.N34S haplotype and noncarrier controls (n = 9) were genotyped for these six 5′ variants, only variant c.-4141G>T was consistently found in p.N34S carriers while absent in controls, indicating that this variant is part of the p.N34S haplotype.
In conclusion, the findings indicate that the p.N34S allele may cause reduced SPINK1 expression. Further studies are needed to verify whether such a negative effect is also manifested in human pancreatic acinar cells. The identification of the c.-4141G>T variant extends the number of published variants within the p.N34S-associated haplotype and represents a potential candidate for the pathogenic variant responsible for reduced SPINK1 expression and pancreatitis risk. Finally, the apparent genetic identity of the PaCa44 and PancTu-I cell lines suggests a common origin.