Interaction of a traditional Chinese Medicine (PHY906) and CPT-11 on the inflammatory process in the tumor microenvironment

Abstract

Background: Traditional Chinese Medicine (TCM) has been used for thousands of years to treat or prevent diseases, including cancer. Good manufacturing practices (GMP) and sophisticated product analysis (PhytomicsQC) to ensure consistency are now available allowing the assessment of its utility. Polychemical Medicines, like TCM, include chemicals with distinct tissue-dependent pharmacodynamic properties that result in tissue-specific bioactivity. Determining the mode of action of these mixtures was previously unsatisfactory; however, information rich RNA microarray technologies now allow for thorough mechanistic studies of the effects complex mixtures. PHY906 is a long used four herb TCM formula employed as an adjuvant to relieve the side effects associated with chemotherapy. Animal studies documented a decrease in global toxicity and an increase in therapeutic effectiveness of chemotherapy when combined with PHY906.

Methods: Using a systems biology approach, we studied tumor tissue to identify reasons for the enhancement of the antitumor effect of CPT-11 (CPT-11) by PHY906 in a well-characterized pre-clinical model; the administration of PHY906 and CPT-11 to female BDF-1 mice bearing subcutaneous Colon 38 tumors.

Results: We observed that 1) individually PHY906 and CPT-11 induce distinct alterations in tumor, liver and spleen; 2) PHY906 alone predominantly induces repression of transcription and immune-suppression in tumors; 3) these effects are reverted in the presence of CPT-11, with prevalent induction of pro-apoptotic and pro-inflammatory pathways that may favor tumor rejection.

Conclusions: PHY906 together with CPT-11 triggers unique changes not activated by each one alone suggesting that the combination creates a unique tissue-specific response.

Figures

Figure 1

Effect of PHY906 and/or CPT-11 on tumor, spleen and liver. (a) Effect of PHY906 and/or CPT-11 on tumor size 72 hours after the initiation of treatment. The p-values refer to unpaired Student t test between groups. (b) Multidimensional scaling (MDS) based on the complete filtered data set including 18,549 transcripts demonstrating in Euclidian space distances among tumor, liver and spleen of animals treated with PBS. (c) Unsupervised, self-organizing clustering based on transcripts that passed for each tissue studied a filter requirement of 80% presence and at least one experiment with a ratio above 4 out of 18,549 that originally passed the less stringent filter (experimental clustering based on Kendal's Tau regression model, tumor = 2, 635 transcripts, spleen = 2,079 transcripts, liver = 1,059). (d) Self-organizing clustering based on 700 genes out of 1,132 tumor genes differentially expressed among the four treatment groups (F test, p-value cutoff < 0.001) that passed a filter of 80% presence and a log ratio ≥ 3 in at least one experiment. Highlighted in yellow is the area of the heat map were enhancement of the CPT-11 effects by the combination CPT-11+PHY906.

Figure 2

Effect of PHY906 and/or CPT-11 on tumor specific canonical pathway with immunologic function. (a) Stacked bar chart summarizing the 23 most affected canonical pathways with immunologic function according to IPA based on 7,348 genes with annotated function differentially expressed (t test cutoff p-value < 0.001) between colon 38 tumors and the combined liver and spleen database in animals treated with PBS. The same canonical pathways are portrayed looking at the effects on the tumor microenvironment of (b) PHY906, CPT-11 (c) and CPT-11+PHY906 (d). Finally, differences between CPT-11+PHY906 compared to CPT-11 alone are shown (e).

Figure 3

Immune network predominantly affected by CPT-11 (the analysis performed at high stringency gene selection; t test p-value cutoff < 0.001).

Figure 4

Immune network predominantly affected by the neoplastic process compared to normal tissues (a) and the effect of PHY906 (b), CPT-11 (c), CPT-11+PHY906 (d) or the differential effect of CPT+PHY906 over CPT-11 alone (e) (the analysis performed at low stringency for gene enrichment; t test p-value cutoff < 0.05).

Figure 5

Macrophage infiltration signatures in tumors treated with PHY906 and/or CPT-11. Macrophage infiltration in colon 38 tumors treated with PBS (a), CPT-11 (b), PHY906 (c) or Ininotecan+PHY906 (d). MCP-1 RNA alteration in tumors induced by various treatments (e)

Figure 6

Quantitative PCR validation of selected genes relevant to the study that were identified as differentially expressed according to transcriptional analysis among different treatment groups. For each transcript, the first graph shows RNA array results followed by qPCR validation results. The Y-axis of the Array data graphs shows relative expression ratio of specific gene in samples relative to the mean of the same gene across PBS control group (REA-Relative Expression in Array). For qPCR validation, the data shown are of the relative expression (RE) of a gene of interest per actin molecule in sample compared to the averaged ratio of the gene of interest to actin for the entire control group (Y-axis). Statistics were Student t-tests (two tailed) performed by GraphaPad software. Error bars represent SD. The four graph bars on the X-axis for each graph represent the different treatment groups.