Rapid growth of a hepatocellular carcinoma and the driving mutations revealed by cell-population genetic analysis of whole-genome data

Abstract

We present the analysis of the evolution of tumors in a case of hepatocellular carcinoma. This case is particularly informative about cancer growth dynamics and the underlying driving mutations. We sampled nine different sections from three tumors and seven more sections from the adjacent nontumor tissues. Selected sections were subjected to exon as well as whole-genome sequencing. Putative somatic mutations were then individually validated across all 9 tumor and 7 nontumor sections. Among the mutations validated, 24 were amino acid changes; in addition, 22 large indels/copy number variants (>1 Mb) were detected. These somatic mutations define four evolutionary lineages among tumor cells. Separate evolution and expansion of these lineages were recent and rapid, each apparently having only one lineage-specific protein-coding mutation. Hence, by using a cell-population genetic definition, this approach identified three coding changes (CCNG1, P62, and an indel/fusion gene) as tumor driver mutations. These three mutations, affecting cell cycle control and apoptosis, are functionally distinct from mutations that accumulated earlier, many of which are involved in inflammation/immunity or cell anchoring. These distinct functions of mutations at different stages may reflect the genetic interactions underlying tumor growth.

Fig. 1.

The scheme of sampling from the HCC liver. The resected portion containing the primary tumor is drawn outside of the liver, as indicted by the dotted lines. From the primary tumor, one large section (T0, >50 mm³) and six small sections (T0–T6, <5 mm³, shown as dots) were taken. Six small sections (N1–N6) were also taken from the adjacent nontumor tissues. The recurrent tumors were detected and operated on 15 mo after the first surgery. The larger recurrent tumor resides in the regenerated portion of the liver and a section (R1, >50 mm³) was taken from it, as well as a section (N0) from the adjacent nontumor tissue. Another section (R2) was taken from the smaller recurrent tumor. For more information, please see Materials and Methods.

Fig. 2.

Detection of large indels on chromosome 5 and 6 from sequence reads. (A) For each chromosome, shown are the minor allele frequency (MAF) at heterozygous site in the non-tumor tissue, N0. Each point represents the sum of 50 consecutive polymorphic sites. Non-tumor tissues do not appear to harbor large indels as the frequencies stay relatively constant across regions. (B) The corresponding frequencies in the R1 section. The contrast is clear since defined regions in R1 show characteristic reductions in MAFs. (C) Read depth is shown; red and green lines denote regions of unusually high or low read depth. There is substantial concordance between B and C in delineating regions of aberration. Since they are built on very different data, the concordance lends confidence to the interpretation of chromosomal indels. Two features are noteworthy as indicated by a red bar (a deletion, D5q) and a red box, respectively. The region marked by the red box has the average read depth but MAFs are aberrant there. A possible interpretation is that, in tetraploids, the two homologs exist in a ratio of 3:1, instead of 2:2.

Fig. 3.

Evolution of the tumors inferred from the data of T0–T6, R1, and R2. The table in the inset shows the presence/absence, indicated by +/− , of each foreground mutation in the tumor sections. (+) denotes presence but at a lower frequency. The table defines the cell lineages. Below the red arrow are mutations accumulated during tumor growth. Red shade denotes tumor cell lineages (labeled as π₀–π₃). The closely related noncancerous cell lineage is labeled π_n. Sample sections, shown in brackets, are written beneath or inside the corresponding cell lineages. M1–M4 and M10 mutations affected amino acid sequences, as shown in Table 1. M5–M9 (□) are silent mutations in intergenic or intronic regions. The deletion Δ5q truncated and fused two genes at the breakpoints. This event is labeled M10. Δ5q also deleted two earlier mutations, M1 and M2. Time is marked by the length of the double arrows on the far right. t1 (=15 mo) is the time between the two surgeries. Among the life time collection of mutations, <5% occurred in the duration of t2. Above the red arrow are background mutations, 188 and 19 of which are silent and nonsynonymous, respectively.