Department of Molecular and Cell Biology, Division of Biochemistry and Molecular Biology, University of California, Berkeley, California, U.S.A.

To the Editor

Xeroderma pigmentosum (XP), a rare autosomal recessive disease characterized by clinical and cellular hypersensitivity to ultraviolet (UV) light (^{Cleaver and Kraemer, 1995}), has been classified into eight genetic groups [A–G, and variant (V)]. XP-E (OMIM no. 278740) was reported to have biochemical heterogeneity regarding a damage specific DNA binding (DDB) activity (^{Kataoka and Fujiwara, 1991};^{Keeney et al, 1992}): some strains lacked DDB activity and were termed Ddb^– XP-E, whereas others had activity and were termed Ddb⁺ XP-E. After reinvestigating three Ddb⁺ XP-E cell strains, we recently reported that each had been misclassified, and based on their phenotypes after UV-irradiation (unscheduled DNA synthesis, recovery of RNA synthesis, and recovery of replicative DNA synthesis in the presence and absence of caffeine) we tentatively reclassified them as XP-F (OMIM no. 278760, XP89TO), XP-V (OMIM no. 278750, XP43TO), and ultraviolet–sensitive syndrome (OMIM no. 600630, XP24KO) strains (^{Itoh et al, 2000a}). We therefore supported the tentative proposal of^{Cleaver et al (1999)} that mutations in the DDB2 gene should be solely responsible for XP-E (^{Itoh et al, 2000a};^{Nichols et al, 2000}).

As the assignment of XP43TO as XP-V was based upon the recovery of replicative DNA synthesis in the presence of caffeine, we have analyzed the XP43TO for mutations in the XPV gene. The XPV gene codes for DNA polymerase (^{Johnson et al, 1999};^{Masutani et al, 1999}), an inducible, damage bypass DNA polymerase (^{Yamada et al 2000}). RT-PCR products from base pairs 110–2410 of the XPV cDNA from XP43TO cells showed a broad band compared with the products from normal Turu cells (RT-PCR1, Figure 1a). This broadening was not seen, however, for products from base pairs 898–2416 (RT-PCR2, Figure 1a). Thirty-one clones from the XP43TO RT-PCR1 band were sequenced Figure 1b. Twelve of these contained one of four types of exon omission, whereas 19 contained one of three types of abnormal splicing events around exon 4. Moreover, exon 4 contained a GT transversion at nucleotide 727 that would generate an E164ter nonsense mutation Figure 1b. By contrast, six independent clones of this region from normal Turu cells had the sequence reported by^{Masutani et al (1999)} and^{Johnson et al (1999)}.

Figure 1.

Analysis of XPV cDNA from XP43TO. (A) RT-PCR was performed on total RNA of XP43TO and normal Turu cells as described (^{Itoh et al, 2000a,b}) and PCR products were resolved on a 0.8% 1X Tris-acetate/EDTA (TAE) agarose gel and stained with ethidium bromide. Primers were S1 [5'-ACTGGACCTCCTAGAAAG (bp 110–129)], S9 [5'-GTCCTGGCAAAACTGGCCTG (bp 898–917)], and AS24 [5'-ATCCTACAGGCAAGCCTGAG (bp 2387–2416)]. RT-PCR1 utilized primers S1 and AS24. RT-PCR2 utilized primers S9 and AS24. M: 1 kb DNA ladder (Life Technologies). (B) The structures of the abnormally spliced XPV cDNA from XP43TO cells deduced from sequence analysis of 31 clones are shown in the context of the genomic structure (^{Yuasa et al, 2000}). Boxes show exons (E); lines show spliced out introns. Arrows or arrowheads indicate a (putative) start codon (ATG) or the GT transversion at nucleotide 727, respectively. Parentheses indicate wild-type exon nucleotides omitted. Also indicated are whether the deletions cause frameshifts or produce an in-frame termination codon(*).

Full figure and legend (39K)

To verify that XP43TO cells carry this mutation in genomic DNA, we performed PCR with genomic DNA and direct sequencing of the region between exon 4 and intron 4 and confirmed the presence of the GT transversion Figure 2a. Significantly, nucleotide 727 is the highly conserved G(-1) at the splice donor site of exon 4 Figure 2b. These results confirm that the patient is either homozygous for the transversion or contained a large deletion of the region in the other allele. In either case, XP43TO should be classified as XP-V.

Figure 2.

Analysis of the XPV gene from XP43TO. (A) Sequence of PCR products. PCR was performed using as template 200 ng genomic DNA from XP43TO or normal Turu cells. Primers were S6 [5'-GTCAGCCTATCTCGGCAGAC (bp 638–657)] and ASI1 [5'-GCAGTATGATGCTATGGAAGTAC] from intron 4 of the XPV gene. PCR products were purified on a 3% 1X TAE agarose gel and sequenced. (B) Schematic representation of the mutation in exon 4.

Full figure and legend (15K)

Of the types of XP43TO cDNA found Figure 1b, types 3, 4, and 5 resulted in frameshifts and type 7 had an in-frame stop codon in the sequence inserted from intron 4. Types 1 and 2 would produce a protein deleted for all of the seven DNA polymerase conserved motifs as identified by^{Yamada et al (2000)} and^{Kannouche et al (2001)}. Type 6 would give rise to a 30 amino-acid deletion between the fourth and fifth motif that could compromise the function or stability of the protein. This large distribution of mRNA types could be due to the transversion, or XP43TO could harbor one or more mutations in the early introns, that could also give rise to splice variants. Alternatively, some or all of the alternative splicing events might occur at a low level in normal cells, but the PCR might not be sensitive enough to detect them in the Turu cells. That is, the distribution might include some minor alternatively spliced forms present in normal cells.

We currently favor the hypothesis that all strains originally reported as Ddb⁺ XP-E by cell fusion were originally misclassified; however, strict proof of this proposal is not possible because some of these strains have been lost and others are available only as late passage samples for which phenotypic analysis is not reliable.