Diurnal rhythms (DR) in gene expression in human oral mucosa: Implications for gender differences in toxicity, response and survival and optimal timing of targeted therapy (Rx)

Reviewer: Christine Hill, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 5, 2007

Presenter: Bjarnason, G.A.
Presenter's Affiliation: Toronto-Sunnybrook Regional Cancer Center, Toronto, ON, Canada.
Type of Session: Scientific


  • Diurnal rhythms (DR) have been suggested to be implicated in tumor growth and response to cancer treatment, and have been observed to play a role in pathophysiology of many other human illnesses, including heart disease and peripheral vasculopathies (Hypertension, 2007).
  • Experimentation in cellular culture versus animal models has demonstrated significant differences in cell cycling, apoptosis, and angiogenesis in in vitro settings, where DR do not contribute, and in vivo settings where they may. Furthermore, mouse models, representing nocturnal rhythms, show substantial differences from dog models, which are diurnal.
  • Cellular rhythmicity, observed to be based on a 24-hour cycle in 10-20% of the rodent genome, is controlled by 'clock' genes, such as Wee1, cMyc, Cyclin D1, and Per1/Per2 (Cell, 2006).
  • Some attempts, such as use of hyperfractionated, accelerated radiation therapy, have been made in modern cancer treatment to target cell cycle timing differences between tumor cells and healthy cells.
  • This study was carried out to evaluate daily whole-genome expression in healthy adult humans in order to assess the effects of DR on cell cycling and other processes with potential relevance to tumor growth and cancer treatment.

Materials and Methods

  • The study population consisted of 10 healthy adult volunteers, 5 male and 5 female.
  • Oral mucosa biopsies were obtained from each subject every 4 hours over 24 hours, for a total of 6 biopsy specimens per research subject.
  • Specimens were subjected to microarray analysis. Genes with significant rhythms were identified using COSOPT, an established wave-fitting algorithm designed for circadian rhythm microarray time series analysis.
  • Gene expression patterns were visualized in GeneSpring GX7.3 and were validated by real-time PCR and analysis of variants.


  • 801 rhythmic genes were identified in male subjects, and 810 in females.
  • 90 rhythmic genes were were common to males and females, including previously-established clock genes. Over 700 genes were identified to exhibit rhythmicity in males and not females and vice-versa.
  • 75 rhythmic transcription factor genes were identified in male subjects, and 67 in female subjects. 28 rhythmic transcription factor genes were common between male and female subjects.
  • 71 rhythmic human cancer genes, defined as gene products involved in signaling pathways known to contribute to tumor growth or be targeted by modern cancer treatments, were identified. Peak expression varied significantly between male and female subjects.
  • The majority of gene expression peaking in males was observed to be 4 AM or 4 PM, and in females was 6 AM or 11 AM.
  • Of note, clock-gene expression profiles were inverted from previously demonstrated rhythms in rodents (Cell, 2006), and the authors note that this finding was expected given the noctural nature of rodent circadian rhythms.

Author's Conclusions

  • Gender-specific DR in gene expression exist, and involve many genes which may be implicated in tumor growth and response to treatment.
  • Gender differences in tumor response to chemotherapy, as well as toxicity have been previously described (JCO 24: 3562, NEJM 353:133), and differences in DR may account in part for these findings.
  • In future trials, the identified rhythmic human cancer genes and the rhythms identified may guide experimental therapies.
  • The observation that human DR are inverted from rodent circadian rhythms may be useful in translating future rodent data in to clinical human trials.

Clinical/Scientific Implications

  • The European Organisation for Research and Treatment of Cancer (EORTC) Chronotherapy Group recently published the results of a multicenter randomized trial examining the effectiveness of chronomodulated infusion of fluorouracil, leucovorin, and oxaliplatin over 4 days versus conventional 2 day delivery of the same drugs for patients with metastatic colorectal cancer (JCO, 2006).
  • Sex was observed to be the most important predictor of survival time: In women, use of chronomodulated therapy increased the risk of earlier death by 38%. In men, the risk of death was decreased by 25% with use of chronomodulated therapy.
  • Additionally, alterations in toxicity, including neutropenia and diarrhea, were observed based on gender.
  • Although the findings reported by the current study do not explain or have direct implications with regards to the described EORTC findings, differences in treatment response and chemotherapy clearly exist between genders.
  • Additionally, significant differences in gene expression patterns between genders not only exist, but appear to be predictable. The potential for these genes to be targeted with cancer treatment certainly exists in the future.
  • This study is limited by its small sample size, and further investigation including more subjects as well as tumor and circadian rhythm response to therapies is certainly warranted.