CarcSeq detection of lorcaserin-induced clonal expansion of Pik3ca H1047R mutants in rat mammary tissue

Abstract

Lorcaserin is a 5-hydroxytryptamine 2C (serotonin) receptor agonist and a nongenotoxic rat carcinogen, which induced mammary tumors in male and female rats in a 2-yr bioassay. Female Sprague Dawley rats were treated by gavage daily with 0, 30, or 100 mg/kg lorcaserin, replicating bioassay dosing but for shorter duration, 12 or 24 wk. To characterize exposure and eliminate possible confounding by a potentially genotoxic degradation product, lorcaserin and N-nitroso-lorcaserin were quantified in dosing solutions, terminal plasma, mammary, and liver samples using ultra-high-performance liquid chromatography-electrospray tandem mass spectrometry. N-nitroso-lorcaserin was not detected, supporting lorcaserin classification as nongenotoxic carcinogen. Mammary DNA samples (n = 6/dose/timepoint) were used to synthesize PCR products from gene segments encompassing hotspot cancer driver mutations, namely regions of Apc, Braf, Egfr, Hras, Kras, Nfe2l2, Pik3ca, Setbp1, Stk11, and Tp53. Mutant fractions (MFs) in the amplicons were quantified by CarcSeq, an error-corrected next-generation sequencing approach. Considering all recovered mutants, no significant differences between lorcaserin dose groups were observed. However, significant dose-responsive increases in Pik3ca H1047R mutation were observed at both timepoints (ANOVA, P < 0.05), with greater numbers of mutants and mutants with greater MFs observed at 24 wk as compared with 12 wk. These observations suggest lorcaserin promotes outgrowth of spontaneously occurring Pik3ca H1047R mutant clones leading to mammary carcinogenesis. Importantly, this work reports approaches to analyze clonal expansion and demonstrates CarcSeq detection of the carcinogenic impact (selective Pik3ca H0147R mutant expansion) of a nongenotoxic carcinogen using a treatment duration as short as 3 months.

Keywords: cancer driver mutation; cancer risk assessment; carcinogenicity testing; mutation; nongenotoxic carcinogen; predictive toxicology.

Fig. 1.

Rat body weight gain over the course of the study. Mean weights of rats treated with 0, 30, and 100 mg/kg lorcaserin are indicated by black squares, green circles, and red triangles, respectively. Error bars indicate the SEM. Level of statistical significance relative to vehicle control (0 mg/kg) are indicated as: P ≤ 0.05, thin black outline; P ≤ 0.01, thick black outline, and P ≤ 0.001, blue outline. The red arrow indicates the point in the study when rats were switched from morning dosing to fasting ∼6.5 h preceding afternoon dosing (after completion of wk 9).

Fig. 2.

Distribution of mutants and their log₁₀ MFs across the 1,077-bp CarcSeq target. For the 24-wk samples, the data from replicate 2 are plotted. Pink arrows point to the position of Pik3ca H1047R hotspot mutants in the panel. Amplicon sequences considered MS (red lines) or NMS (black lines) are indicated below the X-axis.

Fig. 3.

Dose- and time-dependent increase in Pik3ca H1047R MF. Numbers of samples with detectable Pik3ca H1047R MFs, as well as Pik3ca H1047R MF values, increased in rat mammary DNA samples with dose and duration of treatment. In the insert, the Pik3ca H1047R MF scale is expanded for samples treated for 12 wk. The 24-wk values are the averages of the MFs measured in replicates 1 and 2. The boxes depict the 25th, 50th, and 75th percentiles of MF, the whiskers are range bars, and all individual MF measurements are plotted. Kruskal–Wallis ANOVA found significant differences among dose groups in the 12- and 24-wk datasets (P = 0.0053 and P = 0.0063, respectively). For each dataset, Dunn’s multiple comparisons test was used to show Pik3ca H1047R MFs in the 100-mg/kg lorcaserin-treated rats were significantly greater than those of 0 mg/kg control rats (P = 0.0063 and P = 0.0064, respectively).

Fig. 4.

Concordance between replicate CarcSeq measurements of Pik3ca H1047R MF in mammary DNA samples of rats treated for 24 wk. Linear regression and correlation of ACB-of Pik3ca H1047R MFs recovered in replicate analyses are shown (A), along with Bland–Altman analysis (B).

Fig. 5.

Concordance between ACB-PCR and CarcSeq measurements of Pik3ca H1047R MF in rat mammary DNA samples. Fluorescent imaging of the 148-bp ACB-PCR products observed in 2 replicate assays is shown (A), along with an example of a standard curve constructed to interpolate MFs of the unknown samples (B). Linear regression and correlation of ACB-PCR and CarcSeq MFs are shown (C), along with Bland–Altman analysis (D).

Fig. 6.

Greater MFs were recovered in MS than in NMS amplicons. Average MF and MAD in MF were determined for individual rats considering MFs in MS and NMS amplicons, separately. These values were log₁₀-transformed to normalize the data. The average of the log₁₀ MFs (geomean MFs) in the MS and NMS amplicons of each dose group are plotted for 12-wk data (A), and the 24-wk data, replicate 1 (B), replicate 2 (C). The average of the log10 MADs in MF values for the MS and NMS amplicons of each dose group are plotted for 12-wk data (D), and the 24-wk data, replicate 1 (E), replicate 2 (F). Error bars correspond to the SEM. Two-way ANOVA P-values are reported for amplicon type (AT), dose, and their interaction (Int).

Fig. 7.

Normalized mutation frequency shows frequent representation at a subset of trinucleotide targets and differences between MS and NMS targets related to amplicon type but not dose.

Fig. 8.

Relationships between tumor response, MAD, and rat weight. The quotient of the total number of tumors observed in the rats of each bioassay dose group divided by the total number of rats in the dose group was determined and correlated with MAD calculated from the 24-wk MF measurements. Linear regression between the 2-yr bioassay tumor response relative to MAD calculated using: (A) the Pik3ca H1047R MF measurements in the replicate datasets, (B) MAD in MF calculated based on all MFs in MS amplicons (red symbols), or all MFs in NMS amplicons (black symbols) is shown. Rat weight correlated significantly with lorcaserin dose (C) and with individual rat MAD in MF (D). In (C) and (D), values corresponding to the different dose groups are identified by symbol color: black, 0 mg/kg; blue, 30 mg/kg; and red, 100 mg/kg. Error bars indicate SD.