The present study shows that pre-treatment with the proteasome inhibitor bortezomib can decrease packing density in MCC, a tissue culture model for solid tumours, and improve the penetration of other drugs through them. Furthermore, through quantification of the distribution of doxorubicin in tumour xenografts, we show better drug distribution and cytotoxicity in tumours derived from loosely packed as compared to tightly packed sub-lines of HCT-8 human colon carcinoma. Bortezomib was also able to modify the distribution of doxorubicin in the tightly packed HCT-8 xenografts, probably by reducing IFP.
Interactions between tumour cells and components of the ECM can protect solid tumours from toxic stimuli [30
]. Agents that modify or disrupt cell-cell or cell-matrix adhesion have been used to overcome cell adhesion-mediated drug resistance (CAM-DR) in multi-cellular spheroids and xenografts. Hyaluronidase decreased resistance of multi-cellular spheroids to several anticancer drugs including paclitaxel, doxorubicin, and vinblastine [15
], while antibodies against E-cadherin or β1
-integrin were shown to increase drug sensitivity in spheroids or solid tumours [16
]. Our previous studies have shown greater drug penetration and efficacy in MCC derived from colon carcinoma cell lines with a defect in alpha-E-catenin and lack of adherens junctions [3
]. Our analyses of doxorubicin distribution and growth delay of tumour xenografts also show significant differences in drug distrubition in epithelioid and round cell pairs of cell lines, where the penetration of chemotherapeutic agents was greater through MCC derived from loosely packed than the tightly packed sub-lines. The AUC for doxorubicin distribution in the loosely packed HCT-81R1 and Ra xenografts was approximately 2-fold greater than that observed in the tightly-packed HCT-8Ea and HCT-8E11 tumours. Differences in tumour drug distribution as a function of packing density have also been reported by Au et al. (Kuh et al., 1999; Zheng et al., 2001): high cell density was shown to reduce the penetration of doxorubicin and paclitaxel in PC3 and human pharynx FaDu tumours in histoculture and in vivo
To assess the ability of anti-adhesive agents to modify drug penetration, we conducted proof of principle experiments in MCC, using the calcium chelating agent EGTA, which disrupts E-cadherin mediated cell adhesion. Pre-treatment with non-toxic doses of EGTA significantly decreased cellular packing density, and led to improvement in drug penetration in MCC derived from colon carcinoma cell lines.
Bortezomib has been reported to alter the adherence of multiple myeloma cells to ECM proteins and bone marrow stromal cells [36
], and to disrupt cell adhesion in spheroids derived from an ovarian cancer cell line [24
]. Bortezomib was also shown to reduce cell adhesion in squamous cell cancer by down-regulation of the desmosomal cadherin Dsg-2 [37
]. These studies provided the rationale for evaluation of bortezomib as a modifier of cell-cell adhesion and cellular packing density in solid tumours.
In our studies, reduction of cellular packing density in MCC might be due either to cell killing by bortezomib or to reduced cell adhesion (or both); enhanced drug penetration is most likely related to reduced packing density although we cannot exclude some effect due to loss of cells from the surface of the MCC after drug treatment. Interactions between tumour cells and components of the extracellular matrix (ECM) have been shown to protect solid tumours from a number of apoptotic stimuli, and agents that modify or disrupt cell-cell or cell-matrix adhesion have been used successfully to overcome cell adhesion-mediated drug resistance (CAM-DR) in multicellular spheroids and xenografts established from a variety of tumours. Previous studies by Au and colleagues (Kuh et al., 1999; Zheng et al., 2001) have shown increased penetration of paclitaxel and doxorubicin following a low-dose pre-treatment with these drugs, and they attributed this to induction of apoptosis and the subsequent increase in interstitial space. Several studies have reported enhanced sensitivity to chemotherapy after bortezomib treatment, although underlying mechanisms have not been elucidated [17
]. The anticancer drugs chosen in our study ranged from ineffective (5-fluorouracil), minimally cytotoxic (gemcitabine), to moderately cytotoxic (doxorubicin) for HCT-8 cells in culture. Bortezomib pre-treatment either reduced or did not change the cytotoxicity of these drugs in monolayer cultures, while it enhanced cytotoxicity in MCC; this result suggests strongly that the effects of bortezomib to influence sensitivity to other drugs is dependent on cell contact.
Drugs are transported from the circulatory system into the interstitial space both by diffusion and by convection. Elevated interstitial fluid pressure has been shown to limit the efficacy of anticancer drugs by reducing trans-capillary transport and tissue penetration by convection [40
]. Agents that reduce tumour IFP have been shown to improve drug distribution and efficacy in pre-clinical and clinical settings [43
]. Agents that reduce IFP have been shown to enhance the transcapillary transport of low molecular weight tracers in experimental rat colon cancer models and NSCLC xenografts and increase the penetration of monoclonal antibodies in various tumour xenografts [44
]. We observed a correlation between cellular packing density and IFP in xenografts derived from HCT-8 sub-lines. Previous studies have reported that high cell density around blood vessels can lead to elevated IFP. It has also been shown that a reduction in tumour cell density following treatment with paclitaxel (due to the induction of apoptosis) can decrease IFP. We observed no significant reduction in cellular packing density in colon carcinoma xenografts but there was a reduction in IFP at 72
h after bortezomib administration, and this was associated with improvement in doxorubicin penetration. The reduction in IFP after bortezomib treatment cannot be attributed to a reduction in MVD, as bortezomib did not demonstrate a significant anti-angiogenic activity in HCT-8 derived tumor xenografts. Furthermore, tumor xenografts derived from the HCT-8Ea cell line showed lower MVD than the HCT-8Ra tumor xenografts, yet exhibited significantly greater IFP. The xenografts that were evaluated are quite resistant to anticancer drugs. Our data show greater doxorubicin cytotoxicity in the loosely packed HCT-8Ra and HCT-81R1 than the tightly packed HCT8-Ea and HCT8-E11 xenografts. These data are similar to our observations of doxorubicin cytotoxicity and drug penetration using the MCC model and provide further evidence for the role of tumour physiology and drug penetration in drug resistance. In addition, our results further support the use of the MCC model in assessing the role of tumour physiology and architecture in chemotherapeutic resistance. Our findings suggest a trend towards greater growth suppression in tumours treated with bortezomib prior to doxorubicin treatment than those treated with doxorubicin alone, but future studies will benefit from using other drugs to which HCT-8 tumours show greater sensitivity or other tumours that are more sensitive to doxorubicin, in order to assess bortezomib’s potential as a chemo-sensitizer.
In summary, we have provided further evidence for the role of tumour micro-environment in contributing to impaired drug distribution and cytotoxicity in solid tumours. In addition, our studies show that bortezomib can modify the microenvironment and enhance drug penetration in xenografts; its potential to enhance the effects of other anticancer drugs for treatment of solid tumours merits further investigation.