RCCs are resistant to infection by serotype 5 adenoviruses due to lack of coxsakie and adenovirus receptor (CAR) expression.10,24,29,31
To circumvent this problem, we developed a chimeric serotype 5/serotype 3 modified knob adenovirus to deliver mda-7
/IL-24 to RCCs and infect these cells in a CARi-ndependent fashion: Ad.5/3-mda
At low multiplicities of infection, Ad.5/3-mda
-7 but not Ad.5-mda
-7 infected RCCs and promoted expression of MDA-7/IL-24 (). These findings were also reflected in colony formation assay data ().
Figure 1A–D Infection of RCCs with Ad.5/3-mda-7 causes MDA-7/IL-24 secretion into the growth media that has a potent toxic “bystander” effect via CD95 signaling. (A) A498 or UOK121LN cells were cultured for 24 h and then infected with increasing multiplicities (more ...)
MDA-7/IL-24 is a novel cancer-specific cytotoxic cytokine that can be secreted from cells infected with a recombinant adenovirus containing the mda
We next determined whether direct infection of RCCs with Ad.5/3-mda
-7 also permitted MDA-7/IL-24 to be secreted into the culture media and whether this also had a lethal “bystander
” effect on uninfected RCCs. Infection of UOK121LN cells with Ad.5/3-mda
-7, but not Ad.5-mda
-7, increased expression of MDA-7/IL-24 in the culture media of cells (, upper immunoblots). Transfer of the conditioned media containing MDA-7/IL-24, but not “spent” media from vector infected cells, onto uninfected UOK121LN cells increased expression of MDA-7/IL-24. It has been argued that MDA-7/IL-24 can set up a self-stimulatory paracrine loop in tumor cells and in agreement with data in other tumor types, knockdown of MDA-7/IL-24 expression in uninfected cells blocked the conditioned media containing the cytokine from inducing MDA-7/IL-24 expression (, lower immunoblots).38
The lethal effects of Ad.5/3-mda
-7, in infected cells were blocked by knockdown of CD95 expression (). Transfer of media containing MDA-7/IL-24 onto uninfected RCCs promoted cell killing, an effect that was also blocked by knockdown of CD95 or knockdown of MDA-7/IL-24.
Infection of primary renal epithelial cells with Ad.5-mda-7 resulted in expression of MDA-7/IL-24 without causing significant toxicity to these non-transformed cells ( and data not shown). Forty-eight hour after infection, conditioned media from Ad.5-cmv, Ad.5-mda-7 and Ad.5-mda-7 SP- (lacking the signal peptide for cytokine secretion)-infected primary renal epithelial cells was removed and placed onto uninfected UOK121LN carcinoma cells. As a single agent, conditioned media containing from Ad.5-mda-7-infected cells, but not Ad.5-cmv or Ad.5-mda-7 SP-infected cells, suppressed the growth of RCCs as measured in an MTT assay by 38 ± 4% (). In a dose-dependent fashion conditioned media containing MDA-7/IL-24 promoted cell killing in short-term true viability assays ().
Previously we have noted that inactivation of ERK1/2 and AKT and activation of JNK1/2 and p38 MAPK were hallmarks of MDA-7/IL-24 toxicity in renal carcinoma cells.21
In A498 cells molecular inhibition of PI3K, mTOR or MEK1/2 suppressed basal levels of phospho-ERK1/2 and of phospho-(S473)-AKT and in cells infected with Ad.5/3-mda
-7 the dephosphorylation of ERK1/2 and of AKT was enhanced wherein kinase activity was largely abolished (). Knock down of mTOR reduced total and phospho-mTOR levels and MDA-7/IL-24 expression also reduced phospho-mTOR levels, and the combinations of Ad.5/3-mda
-7 and molecular inhibition of PI3K, mTOR or MEK1/2 further reduced mTOR phosphorylation.
Figure 2 Ad.5/3-mda-7 lethality is enhanced by combined molecular inhibition of PI3K/MEK/mTOR pathways in A498 cells. (A) A498 cells were infected with empty vector control virus (Ad.5/3-cmv) or with viruses expressing MDA-7/IL-24 (Ad.5/3-mda-7); and in parallel (more ...)
Expression of MDA-7/IL-24 enhanced JNK1/2 phosphorylation that was further increased by combined inhibition of signaling by: PI3K + MEK1/2; mTOR + MEK1/2; PI3K + mTOR; or PI3K + MEK1/2 + mTOR (). This correlated with increased expression of the toxic BH3 domain protein BAX and decreased expression of MCL–1. Prior studies from our group demonstrated that MDA-7/IL-24-induced PKR-like endoplasmic reticulum kinase (PERK) signaling was essential for cytokine lethality.21,27
MDA-7/IL-24 enhanced PERK and eIF2α phosphorylation that was strongly enhanced by combined inhibition of PI3K + MEK1/2; mTOR + MEK1/2; PI3K + mTOR; or PI3K + MEK1/2 + mTOR. These findings correlated with decreased expression of the PERK chaperone BiP/GRP78.
We next determined the impact of inhibiting signaling pathways on Ad.5/3-mda-7 toxicity in A498 RCCs. Molecular inhibition of mTOR; PI3K; MEK1/2; PI3K + MEK1/2; mTOR + MEK1/2; PI3K + mTOR; or PI3K + MEK1/2 + mTOR caused modest increases in the basal level of apoptosis in Ad.5/3-cmv infected cells (). Combined inhibition of PI3K + MEK1/2 or PI3K + mTOR modestly enhanced MDA-7/IL-24 toxicity whereas combined PI3K + MEK1/2 + mTOR inhibition strongly enhanced cell killing. The short-term viability data was also reflected in assays determining total cell numbers (). However, in long-term colony formation assays only combined PI3K + MEK1/2 + mTOR inhibition strongly enhanced MDA-7/IL-24 toxicity (). In UOK121LN RCCs only combined molecular inhibition of PI3K + MEK1/2 + mTOR signaling significantly enhanced MDA-7/IL-24 toxicity and suppression of growth in short term assays whereas molecular inhibition of any individual signaling pathway promoted MDA-7/IL-24 toxicity in long-term colony formation ().
Figure 3 Ad.5/3-mda-7 lethality is enhanced by combined molecular inhibition of PI3K/MEK/mTOR pathways in UOK121LN cells. (A and B) UOK121LN cells were infected with empty vector control virus (Ad.5/3-cmv) or with viruses expressing MDA-7/IL-24 (Ad.5/3-mda-7); (more ...)
Multiple drugs have been developed for clinical application that are designed to inhibit the PI3K/MEK1/2/mTOR signaling pathways and the multi-kinase inhibitor sorafenib which suppresses ERK1/2 pathway activity as well as inhibiting receptor tyrosine kinases is an FDA approved therapeutic for RCC. We next examined the impact of the PI3K inhibitor PX-866; the MEK1/2 inhibitor PD184352; and the mTOR inhibitors rapamycin and PI-103 on MDA-7/IL-24 lethality in RCCs. Use of kinase inhibitor drugs, as compared to transfecting plasmids to express dominant negative inhibitory proteins, resulted in a greater suppression of ERK1/2 and AKT signaling under basal conditions and promoted a greater level of PERK/eIF2α and JNK1/2 activation ().37
Inhibition of MEK1/2 and PI3K significantly enhanced MDA-7/IL-24 lethality in RCCs, compared to inhibition of MEK1/2 and mTOR ( and C
). Combined inhibition of mTOR, PI3K and MEK1/2 enhanced MDA-7/IL-24 toxicity to a greater extent than inhibition of MEK1/2 and PI3K (). Sorafenib enhanced MDA-7/IL-24 toxicity ().
Figure 4 Inhibition of mTOR, PI3K and MEK1/2 signaling using clinically relevant small molecule kinase inhibitors enhances Ad.5/3-mda-7 lethality. (A) A498 cells were infected with empty adenovirus vector (Ad.5/3-cmv) or an adenovirus vector expressing MDA-7/IL-24 (more ...)
In general agreement with data in and C; , in colony formation assays combined inhibition of mTOR, PI3K and MEK1/2 or sorafenib treatment enhanced MDA-7/IL-24 toxicity to a greater extent than inhibition of MEK1/2 and PI3K or of mTOR (). Using purified GST-MDA-7, we noted that sorafenib synergized in median dose effect colony formation assays with the cytokine to kill RCCs with combination index (CI) values of less than 1.00 ().
GST-MDA-7 and sorafenib synergize to kill RCCs
Based on our data in –
we next determined the signaling and apoptosis pathways that were being modulated in RCCs. Expression of activated AKT and activated MEK1 significantly protected cells from inhibition of mTOR, PI3K and MEK1/2 to a similar extent as expression of activated AKT, activated MEK1 and activated mTOR (, p < 0.05). Inhibition of JNK1/2 signaling blunted MDA-7/IL-24 lethality and it suppressed the elevation of MDA-7/IL-24 toxicity by combined inhibition of mTOR, PI3K and MEK1/2.37
Figure 5 Inhibition of signaling pathways enhances Ad.5-mda-7-induced CD95 activation and facilitates killing independently of CD95 activation. (A) A498 cells were infected with Ad.5/3-cmv or Ad.5/3-mda-7 and in parallel transfected with plasmids to express activated (more ...)
Prior studies in RCCs have shown that MDA-7/IL-24 kills primarily via activation of CD95 (the extrinsic pathway), and our present studies also demonstrated that secreted MDA-7/IL-24 killed bystander RCCs in a CD95 dependent fashion (). The ability of MEK1/2 and PI3K inhibition to enhance MDA-7/IL-24 toxicity was suppressed by ~50% when extrinsic apoptosis pathway signaling was blocked (). However, when signaling via the intrinsic pathway was blocked, MDA-7/IL-24 toxicity was abolished. As MDA-7/IL-24 activates CD95 in RCCs we determined whether modulation of signaling pathway functions altered the levels of MDA-7/IL-24-induced CD95 activation. Use of molecular tools modestly enhanced MDA-7/IL-24-induced CD95 surface localization and DISC formation whereas use of small molecule kinase inhibitors, that inhibited the signaling pathways to a greater extent, caused a much larger enhancement of CD95 surface localization, that in the case of PX-866 + PD184352 treatment or sorafenib treatment, was greater than additive (). These findings using kinase inhibitors correlated with increased association of pro-caspase 8 and FADD with CD95 (, upper blots).
Based on the data in , we wished to determine whether the in vitro toxic “bystander” effect of secreted MDA-7/IL-24 on uninfected tumor cells could be translated into an in vivo model system. A498 RCC tumors were established on two flanks of an athymic mouse. One tumor was injected every second day with Ad.5/3-cmv
or with Ad.5/3-mda
-7 (total two injections) and the volumes of the tumors on both sides of the animal measured using calipers. Infection of tumors with Ad.5/3-mda
-7, but not Ad.5/3-cmv
, caused a rapid suppression of the tumor growth rate and significantly increased animal survival ( and B
(p < 0.05)). Infection of animals with Ad.5/3-mda
-7, but not Ad.5/3-cmv
, also suppressed the growth rate of uninfected tumors growing on the opposite flank, as previously observed in the context of prostate tumors implanted on opposite flanks of nude mice.34
This correlated with caspase 3 cleavage; MDA-7/IL-24 expression and increased apoptosis in both the Ad.5/3-mda
-7-infected and uninfected “bystander” RCC tumors ( and D
). Of particular note, Ad.5/3-cmv
virus infection promoted recruitment of CD11c and CD11b positive immune cells (dendritic and natural killer cells) into tumors, as judged by CD11c and CD11b staining in Ad.5/3-cmv
infected tumors but not in the contra-lateral uninfected “bystander” tumor. In contrast to empty vector virus, Ad.5/3-mda
-7 infection resulted in CD11c and CD11b staining in both the infected tumor and in the contra-lateral uninfected bystander tumor. Together with our in vitro data, these findings argue that MDA-7/IL-24 secreted from a renal carcinoma tumor in vivo has antitumor activity on an uninfected distant RCC tumor.
Figure 6 Ad.5/3-mda-7 infection of RCC tumors in vivo exhibits a toxic bystander effect on uninfected RCC tumors. (A) A498 cells were injected into the rear right and left flanks of athymic mice. Tumors of ~150 mm3 grew over the following 29 days. Animals (more ...)
Finally, based on our in vitro studies we determined whether the in vivo RCC antitumor effects of MDA-7/IL-24 could be enhanced by the clinically relevant RCC therapeutic sorafenib.38,39
Pre-formed A498 tumors (~150 mm3
) were injected with either Ad.5/3-cmv
. A second adenoviral infusion was performed 48 h after the first infection. Twenty-four hour after the first viral infection animals were treated with vehicle or sorafenib (20 mg/kg) for the following 4 days. Infection with Ad.5/3-mda-7
or treatment with sorafenib reduced the rate of tumor growth (). However, combined exposure to Ad.5/3-mda-7
and sorafenib promoted a significantly greater suppression of tumor growth than either agent individually. Based on animal welfare guidelines, animals carrying tumors > 2.0 cm3
in volume are humanely sacrificed; treatment of animals with empty vector virus and sorafenib modestly increased animal survival whereas treatment with Ad.5/3-mda-7
significantly increased survival (). Combined treatment of animals with sorafenib and Ad.5/3-mda
-7 enhanced animal survival beyond that caused by administration of Ad.5/3-mda
-7 alone (p < 0.05). In isolated tumors it was noted that cleavage of pro-caspase 3 and TUNEL positivity was increased by Ad.5/3-mda-7
+ sorafenib treatment and decreased Ki67 reactivity ( and D
). Infection with Ad.5/3-mda
-7 also promoted recruitment of CD11b and CD11c positive immune cells into the tumor; immune cells including monocytes, macrophages/dendritic cells, neutrophils, natural killer cells and B cells. In addition we found that: (a) Ad.5/3-mda
-7 and sorafenib did not interact to kill primary human renal epithelial cells in vitro and; (b) no obvious normal tissue toxicity by H&E staining was noted in to be induced in either the kidneys or livers of animals exposed to Ad.5/3-MDA-7 and sorafenib (data not shown). Collectively, our data argue that MDA-7/IL-24 augments the antitumor effects of sorafenib against established renal carcinoma tumors.
Figure 7 Sorafenib enhances the lethal effects of Ad.5/3-mda-7 in vivo. (A) A498 tumors formed over 29 days in the flank of athymic mice (~150 mm3). Animals were separated into tumor volumes of approximate equivalent mean tumor size and standard error. (more ...)