Characterisation of the HLA-DRB1*07:01 biomarker for lapatinib-induced liver toxicity during treatment of early-stage breast cancer patients with lapatinib in combination with trastuzumab and/or taxanes.

HLA-DRB1*07:01 allele carriage was characterised as a risk biomarker for lapatinib-induced liver injury in a large global study evaluating lapatinib, alone and in combination with trastuzumab and taxanes, as adjuvant therapy for advanced breast cancer (adjuvant lapatinib and/or trastuzumab treatment optimisation). HLA-DRB1*07:01 carriage was associated with serum alanine aminotransferase (ALT) elevations in lapatinib-treated patients (odds ratio 6.5, P=3 × 10-26, n=4482) and the risk and severity of ALT elevation for lapatinib-treated patients was higher in homozygous than heterozygous HLA-DRB1*07:01 genotype carriers. A higher ALT case incidence plus weaker HLA association observed during concurrent administration of lapatinib and taxane suggested a subset of liver injury in this combination group that was HLA-DRB1*07:01 independent. Furthermore, the incidence of ALT elevation demonstrated an expected correlation with geographic HLA-DRB1*07:01 carriage frequency. Robust ALT elevation risk estimates for HLA-DRB1*07:01 may support causality discrimination and safety risk management during the use of lapatinib combination therapy for the treatment of metastatic breast cancer.

Comprehensive genome-wide evaluation of lapatinib-induced liver injury yields a single genetic signal centered on known risk allele HLA-DRB1*07:01.

Lapatinib is associated with a low incidence of serious liver injury. Previous investigations have identified and confirmed the Class II allele HLA-DRB1*07:01 to be strongly associated with lapatinib-induced liver injury; however, the moderate positive predictive value limits its clinical utility. To assess whether additional genetic variants located within the major histocompatibility complex locus or elsewhere in the genome may influence lapatinib-induced liver injury risk, and potentially lead to a genetic association with improved predictive qualities, we have taken two approaches: a genome-wide association study and a whole-genome sequencing study. This evaluation did not reveal additional associations other than the previously identified association for HLA-DRB1*07:01. The present study represents the most comprehensive genetic evaluation of drug-induced liver injury (DILI) or hypersensitivity, and suggests that investigation of possible human leukocyte antigen associations with DILI and other hypersensitivities represents an important first step in understanding the mechanism of these events.

Defective bypass replication of a leading strand cyclobutane thymine dimer in xeroderma pigmentosum variant cell extracts.

Xeroderma pigmentosum variant (XP-V) is an inherited disorder resulting in hypersensitivity to the cytotoxic, mutagenic, and carcinogenic effects of UV light. There is evidence suggesting that XP-V cells carry a defect in the replication of UV-induced DNA damage, leading to mutations in genes, e.g., proto-oncogenes and tumor suppressor genes, of exposed skin cells. Using an in vitro assay to quantitatively evaluate replication of the most prevalent UV-derived DNA lesion, the cis,syn-thymine dimer (T x T), we have recently found that a T x T located on the leading strand can be bypassed by a bona fide human replication fork but can also induce fork uncoupling with selective synthesis of the undamaged lagging strand (D. Svoboda and J-M. Vos, Proc. Natl. Acad. Sci. USA, 92: 11975-11979, 1995). We now report the application and further refinement of this sensitive assay to the replication of a T x T-containing template by XP-V cell-free extracts. In comparison to normal controls, a 10-26-fold deficiency in the bypass replication of T x T was observed in XP-V cell extracts. In contrast, the disease extracts were as competent as controls for replication of the undamaged TT plasmid and for leading T x T-induced fork uncoupling. Besides mismatch repair and nucleotide excision repair, the bypass replication defect of XP-V may represent a novel category of hereditary mutator phenotypes affecting DNA damage processing.

Reparative strand incision in saponin-permeabilized human fibroblasts.

The damage-directed strand incision step in the nucleotidyl DNA excision-repair pathway (NDERP) was characterized in quiescent monolayer cultures of human fibroblasts in which the plasma membrane was selectively permeabilized with saponin. When permeable normal human fibroblasts (NHF) were incubated in a DNA-repair assay mixture lacking the deoxyribonucleoside triphosphate precursors, the numbers of UV-dependent DNA-strand breaks were increased by about 9-fold consistent with the uncoupling of incision from gap-filling DNA synthesis and ligation. In uncoupled NHF omission of ATP reduced the numbers of UV-dependent strand breaks by 84% confirming the requirement for ATP for reparative strand incision. Time-course experiments indicated that the maximum rate of strand incision occurred in the first 10 min of incubation of permeable cells and diminished to 16-28% of this rate between 30 and 60 min of incubation. The initial rate of incision in permeable NHF was estimated to be 20% of that seen in intact fibroblasts. Dose-response studies indicated an initial saturation of strand incision activity at fluences between 10 and 25 J/m2. In permeable group A xeroderma pigmentosum fibroblasts (XPA) few UV-dependent incisions were produced after 10-25 J/m2. In the xeroderma pigmentosum variant (XPV) strain that we studied, strand incisions saturated at a plateau level that was about twice that seen in the NHF strain suggesting the preservation of a higher level of incision activity after permeabilization. After fluences above 50 J/m2 additional strand incision was observed in all cell strains reflecting the activity of a damage-dependent endodeoxyribonuclease that is independent of the NDERP. Saponin-treated fibroblasts were also permeable to pancreatic deoxyribonuclease I and the UV-DNA endonuclease from M. luteus indicating that these preparations may be used for in vitro complementation.

Formation of DNA strand breaks in UV-irradiated human fibroblast preparations.

The formation of DNA strand breaks was characterized in human fibroblasts prepared by several methods. In quiescent monolayer cultures of normal human fibroblasts (NHF), exposure to 254 nm radiation (UV) caused the rapid appearance of DNA strand breaks as monitored by alkaline elution analysis. Maximal levels of DNA breaks were seen 30 min after 10 J/m2; thereafter, strand breaks disappeared. Breakage soon after irradiation appeared to saturate at fluences above 10 J/m2. Xeroderma pigmentosum fibroblasts belonging to complementation group A (XPA) did not display this response which reflects operations of the nucleotidyl DNA excision repair pathway. When fibroblast strains were released from culture dishes by enzymatic digestion with trypsin or by scraping with a rubber policeman, UV-dependent DNA breakage displayed altered dose and time responses. Few breaks were detected in detached preparations of NHF after 10 J/m2 indicating inactivation of nucleotidyl DNA excision repair. The fluence response in detached fibroblasts was linear up to an incident fluence of 100 J/m2. Moreover, after 25 or 50 J/m2, strand breaks accumulated as a linear function of time for up to 2 h after irradiation. This UV-dependent and time-dependent incision activity was also observed in XPA monolayers and released-cell preparations. In permeable fibroblast preparations, DNA breaks accumulated in unirradiated cells that had been released with trypsin or by scraping. Permeabilization in situ saponin to open the plasma membrane produced a cell preparation that accumulated fewer UV-independent breaks. In saponin-permeabilized NHF that were irradiated with 10 J/m2, UV-dependent strand incision activity occurred at about 30% of the rate of incision seen in intact monolayer NHF. These results reveal at least 3 DNA strand incision activities in human fibroblast preparations of which only one reflects operation of the nucleotidyl DNA excision repair pathway.

Biochemical characteristics of endonuclease activity within the nucleotidyl DNA excision repair pathway of permeable human fibroblasts.

DNA strand breage in response to damage produced by UV (254 nm) radiation was characterized after permeabilization of diploid normal and xeroderma pigmentosum variant fibroblasts. The breakage reaction required ATP, Mg2+ and sucrose for maximal activity and was inhibited by 150 mM Na+ or K+ and 1 mM N-ethylmaleimide. ATP-dependent strand breakage was saturated at UV fluences of above 10 J/m2 and in the presence of DNA precursors breakage was rapidly followed by DNA polymerase and ligase activities to seal the strand breaks. The biochemical features of strand breakage in irradiated permeable cells suggest an enzymatic process. These results, therefore, provide an indication of the biochemical requirements for the rate-limiting strand incision step within the nucleotidyl DNA excision repair pathway.

Analysis of a clonal selection event during transposon-mediated nested-deletion formation in rare BAC and PAC clones.

Nested deletions from one end of the genomic DNA in bacterial artificial chromosomes (BACs) and P1 artificial chromosomes (PACs) are readily generated by inserting a loxP site-containing Tn10 minitransposon into the recombinant clone and transducing with P1 phage. Although the size of clones in the deletion series is largely random, in about 5% of BACs and PACs the distribution appears skewed to a certain length, and in rare cases (<1%) is definitely skewed to a particular size. Here we investigate this relatively rare phenomenon and validate that sequence-specific transposon insertions are not the cause of such skewed nested-deletion libraries. Instead, a detailed analysis of our experiments with a BAC clone demonstrating this unusual feature indicates that deletions of a certain size arise from clonal expansion of a transposon insertion as a result of transient derepression of the transposase gene prior to IPTG induction. Transposition itself shows no bias to any particular region of insert DNA in the clone. We suggest a simple modification to the procedure for generating nested-deletions that allows all BACs and PACs to produce nested-deletions of random size. These findings should provide additional insight into the causes of site selectivity in genomic clones with other inducible transposon systems.