This is the first investigation demonstrating that SMO expression in BC tissues is significantly lower than in nonneoplastic tissues, by RT-PCR and enzyme activity analyses.
It has been proposed that the H2O2 produced during the oxidation of PA by SMO may contribute to the level of apoptosis in BC cell lines [11, 15]. Thus, it is tempting to speculate that the significant decreases in SMO activity that we observed in BC tissues may contribute to tumor growth through a decreased rate of endogenous apoptosis resulting from decreases in the local concentrations of H2O2. SMO dysregulation has also been observed in prostate cancer, by using image analysis techniques and TMAJ software tools , and in ulcerative colitis, by TaqMan-based real-time PCR . In both cases the level of SMO expression was observed to be upregulated. To explain these apparently contradictory results of SMO expression the process of carcinogenesis has to be considered. It is well known that a large number of human cancer types has been directly associated to chronic inflammation, a temporally limited adaptive response. During inflammation there is an oxidative stress and many evidences point out that SMO activity participates in producing reactive oxygen species (H2O2) [6, 36, 37]. When the inflammation-driven tumour progressively develops, the H2O2 produced by oxidation of PA by SMO and APAO may potentially negatively contribute to cell proliferation. In this scenario, we expected to observe an inversion tendency of SMO and APAO gene expression, resulting in a lower ROS production that no longer contrasts the tumour progression.
In line with this hypothesis, we confirm that APAO enzyme activity was significantly lower in BC tissue , contributing to a reduction of the cellular H2O2 level. We also confirm the results of previous studies on SSAT and ODC enzyme activities carried out in human BC tissues compared with nonmalignant control, indicating a high ODC and SSAT activities in BC samples [12, 38, 39]. Manni et al.  reported that increased ODC activity is associated with an increased risk of both disease recurrence and death, while Deng et al.  showed that the overexpression of ODC in BC tissues correlates with the TNM grading system. Wallace et al.  reported that the decrease of APAO activity in BC tissues positively correlated with the aggressiveness of the tumor, while the increase of SSAT activity showed a tendency to be indicative of a poor prognosis. It was proposed that SSAT and APAO interplay could produce an efficient system to generate locally high concentrations of H2O2 that drives cells to the death signaling pathway, notwithstanding SMO activity was not considered in this context .
In BC cells this normal death-generating pathway is blocked by the decrease in activity of APAO and, thus, acetylPA accumulate within the tissue. AcetylPA, unlike PA, are not significantly toxic to tumor cells and may, thus, remain within the tumor cells without causing any negative effects . On the other hand, the increase in SSAT activity explains the high level of N1, N12-diacetylspermine (DiAcSpm) in the urine of BC patients, which proved to be a more sensitive marker than CEA, CA19-9 and CA15-3 in BC at early stages .
Considering this altered PA homeostasis in BC tissue, a low gene expression of both SMO and APAO is in line with the tumoral blocked death-generating pathway and the high proliferating cellular rate. BENSpm and CPENSpm have been successfully employed as antiproliferate compounds on some human BC cell lines [11, 15, 17, 18] but in Phases I and II of clinical trials gave poor positive outcomes [14, 18, 20–22]. The H2O2 produced through BENSpm-induced PA catabolism was found to be derived exclusively from SMO and not through APAO activity. This data suggested that SMO activity is the major mediator of the cellular response of BC cells to BENSpm and that APAO plays little or no role in this response . It has been shown that the utilization of CPENSpm has produced overlapping results with the BENSpm treatments . We modeled the structure of the complexes formed by the mSMO enzyme with BENSpm and CPENSpm. This modeling analysis, a fast and economic way of screening a large number of potential Spm analogues, has revealed the structural bases of the higher affinity that CPENSpm displays for mSMO active site with respect to BENSpm. The cyclopropyl substituent present in CPENSpm nicely fits in a hydrophobic pocket present in mSMO active site. This interaction, estimated to increase CPENSpm affinity for mSMO, is absent in the modeled BENSpm-mSMO. To confirm that these molecules impair mSMO activity, the inhibition of this enzyme activity by BENSpm and CPENSpm was analyzed. Values of K
i for mSMO competitive inhibition by these two Spm analogues indicate that they can be considered good in vitro inhibitors of SMO activity, with CPENSpm being more reactive than BENSpm. These results are apparently in contradiction with previous data reported by Wang et al. , indicating that both BENSpm and CPENSpm molecules are poor inhibitors of the human SMO enzyme. However, in that work the authors were searching for strong SMO inhibitors and used a very low inhibitor concentration (10 μM) compared to substrate concentration (250 μM). The high Spm concentration utilized in Wang's experiments would explain the poor inhibition observed . In our experimental conditions the BENSpm and CPENSpm concentration was in the range of 10-4 M in the presence of 5-10 × 10-6 M Spm. BENSpm and CPENSpm show K
i values (3.8 × 10-4 M and 8.5 × 10-5 M, respectively) comparable to that of MDL72527 (6.3 × 10-5 M), which can be considered a good inhibitor of SMO activity . The finding that both BENSpm and CPENSpm are inhibitors of the SMO catalytic activity could explain their SMO induction effect as a cellular mechanism to overwhelm enzyme inhibition. This novel finding on the inhibitory properties of BENSpm and CPENSpm should not be underestimated and could explain the clinical trials failure of BENSpm. Nevertheless, a key question is to understand how intracellular SMO and SSAT up-regulations are exerted by BENSpm and CPENSpm treatment. One hypothesis is that these analogues compete with natural PA for uptake when using the PA transporter to gain entry into the cell .