In Silico cancer cell versus stroma cellularity index computed from species-specific human and mouse transcriptome of xenograft models: towards accurate stroma targeting therapy assessment

Abstract

Background: The current state of the art for measuring stromal response to targeted therapy requires burdensome and rate limiting quantitative histology. Transcriptome measures are increasingly affordable and provide an opportunity for developing a stromal versus cancer ratio in xenograft models. In these models, human cancer cells are transplanted into mouse host tissues (stroma) and together coevolve into a tumour microenvironment. However, profiling the mouse or human component separately remains problematic. Indeed, laser capture microdissection is labour intensive. Moreover, gene expression using commercial microarrays introduces significant and underreported cross-species hybridization errors that are commonly overlooked by biologists.

Method: We developed a customized dual-species array, H&M array, and performed cross-species and species-specific hybridization measurements. We validated a new methodology for establishing the stroma vs cancer ratio using transcriptomic data.

Results: In the biological validation of the H&M array, cross-species hybridization of human and mouse probes was significantly reduced (4.5 and 9.4 fold reduction, respectively; p < 2x10-16 for both, Mann-Whitney test). We confirmed the capability of the H&M array to determine the stromal to cancer cells ratio based on the estimation of cellularity index of mouse/human mRNA content in vitro. This new metrics enable to investigate more efficiently the stroma-cancer cell interactions (e.g. cellularity) bypassing labour intensive requirement and biases of laser capture microdissection.

Conclusion: These results provide the initial evidence of improved and cost-efficient analytics for the investigation of cancer cell microenvironment, using species-specificity arrays specifically designed for xenografts models.

Figure 1

Schematic diagram for the biological and computational design and annotation of the H&M microarray. The whole array design and analysis includes three stages, as described in the Materials and Methods Section. There were 3.6k newly designed probes in step A and 7.1k newly designed probes in step B. All probes on H&M (Agilent GEO platform GPL10749) were optimally and extensively annotated for cross-species hybridization as shown in Dataset S1. The Agilent pre-annotated cross-species hybridization probes from step A are given in Datasets S2-3, and the processes used to design the new probes in version 1 were previously published [17].

Figure 2

Comparison of expression of custom designed species-specific probes (Y axis) against the Commercial Agilent 44k array probes (X axis). The log2 expression of custom species-specific probes (Y axis) was plotted against that of their Agilent counterpart targeting the same gene (mean expression value, n = 3 arrays). The dashed diagonal corresponds to equal expression between the two types of probes. The density distribution of the number of pairs of probes is shown as hexagons with increased color intensity. As shown in Panels A and B, the human custom species-specific probes obtain a lower expression than their Agilent counterpart with mouse RNA and the similar results for mouse probes (4.5 median fold reduction, 95% CI: 4.2, 4.9 for human probes; and 9.4 median fold reduction, 95% CI: 8.8, 10.3 for mouse probes; P < 2.2x10^-16 for both; Non-parametric Mann-Whitney U tests). Panels C and D show that these species-specific probes of the H&M array perform as well on average as those Agilent probes when exposed to the intended RNAs. Human-mouse gene cross-hybridization patterns we discovered are provided in Dataset S1 (http://lussierlab.org/publications/HsMm_array).

Figure 3

Species-specific probes of pair-wise human and mouse homologous housekeeping genes. The expression of 26 species specific probes (SSPs) corresponding to 13 pairs of human and mouse homologous housekeeping genes (HK) are shown according to two experimental conditions: exposure to the H&M array to Human RNA (X axis) or mouse RNA (Y axis). Each point is the mean expression value of 3 arrays and the whiskers are the 95% confidential intervals (n = 6 arrays total). A median fold change for the 13 pairs of SSPs resulted in a 10.4 fold change for human SSPs (panel A) and 9.7 fold change for mouse SSPs (panel B). Homologous genes utilised: RNASEH1, PMPCB, SFRS8, PRDX6, TRAPPC4, MATR3, NIPA2, MRPL49, NOL7, VPS26A, HNRPDL, RPL39, OSBP.

Figure 4

Validation of cancer vs stromal RNA ratio using the expression of pair-wise human and mouse homologous housekeeping genes measured by species-specific probes. The log2 expression ratio between human SSPs and mouse SSPs were plotted in three cases using the initial experimental data, when probes were exposed to: A) only human RNAs, B) only mouse RNA, and C) both human and mouse RNA. Consistent with the results shown in Figure 3, the expression ratios were larger than 10 (log2(10)=3.3, the dashed blue lines) when SSPs were exposed to the indented RNA only, while the 75th quantile of expression ratios were between 0 and 10 when exposed to equally human and mouse RNAs (panel C).