Although aneuploidy is a global genomic abnormality present in most human cancers, the clonal selection model best explains the action of select activating mutations in oncogenes and homozygous losses of tumor-suppressor genes. Simple gene dosage changes are difficult however, to incorporate into this model, in part due to negative feedback loops that govern major cancer mutational targets (e.g., TP53, PTCH, SMAD4) and essentially preclude a haploinsufficient phenotype. The 17p conundrum may offer a clue to reconciling this difficulty: In comparison to the moderate mutation rate of TP53, many tumors have a disproportionately high frequency of loss of 17p. This discrepancy, and similar discrepancies at other sites of LOH, has long been thought to be due to the presence of undiscovered yet frequently mutated tumor-suppressor genes. However, over 15 years of searching for this grail has distributed bountiful disappointment. It is perhaps time to seriously consider an alternative explanation. Located on 17p adjacent to the TP53 gene, MKK4 is one of the most consistently mutated genes across tumor types, and is located on one of the most frequently lost arms in the human genome. We theorized that a gene dosage-dependent phenotype of MKK4 could plausibly promote the emergence of 17p LOH and thereby the probability of evolving the biallelic inactivation of TP53. Using MKK4 somatic human knockout cancer cells, we observed the proof-of-principle in the downstream gene dosage-dependent phenotypes: heterozygous and homozygous knockouts were progressively deficient in Mkk4 protein, in stress-induced phosphorylation of Jnk, and the resultant upregulation of JUN mRNA. These observations highlight a lack of compensatory regulation when gene dosage changes perturb the Jnk-Jun relationship. Consideration of gene dosage changes specifically affecting members of positive feedback loops may permit integration of the aneuploidy process into a conventional model of clonal selection in tumorigenesis.