We first verified the cytotoxic activity of SE and SP. For this, pooled SE or SP derived from 50 donors was titrated on TZM-bl cells, a HeLa cell derivative widely used in AIDS research that is engineered to express the HIV-1 receptors CD4 and CCR5.32 TZM-bl cells were seeded at a low cell density and treated with increasing SE or SP concentrations in DMEM supplemented with 10% (v/v) heat-inactivated FCS. After two days of incubation at 37°C, cell viability was determined using a colorimetric MTT assay (Fig. 1a) and light microscopy (data not shown). Microscopic evaluation revealed that cells treated with SE as well as SP concentrations as low as ~ 1% already exhibited a spherical shape, which became more evident with higher concentrations of the body fluids. At concentrations of 10% and higher, the cells detached completely from the bottom of the cell culture plate. The MTT-assay measures the conversion of a soluble MTT salt into a purple-colored formazan product in viable cells.33 The MTT assay confirmed these observations and showed a concentration-dependent decrease in viable cells (Fig. 1a), as previously shown.30,31 SE or SP concentrations > 1% led to an almost complete loss of viable cells and concentrations as low as 0.3% already resulted in a 20% reduction of viability (Fig. 1a). Thus, SE as well as SP are highly cytotoxic under standard cell culture conditions involving utilization of 10% FCS, confirming previous results of others and us. 6–14,17−19
Next, we determined the cytotoxic activity of synthetic spermine and spermidine using the same experimental setup as described above. MTT assays performed after 2 days showed that in the presence of 10% FCS, concentrations of ≥ 10 µM of both polyamines were strongly cytotoxic for TZM-bl cells (Fig. 1b), confirming previous data obtained with several cell lines such as A549 lung adenocarcinoma and HCT116 colon adenocarcinoma cells,30,31 human fibroblasts,24 rat neurons,34 or murine lymphocytes.11 To quantify the polyamine concentrations in the pooled SP sample, liquid chromatography-tandem mass spectrometry was applied,35 revealing concentrations of ~ 3 mM spermine, 155 µM spermidine, 80 µM putrescine and 255 µM L-ornithine (Fig. S1), in line with published data.20 Thus, the spermine concentrations in SP are in a range that can explain the cytotoxicity of SP and suggests that spermine is the major contributor to the cell-damaging activity in semen.
To further investigate the role of serum in the observed effects, we titrated FCS on TZM-bl cells supplemented with PBS, 100 µM spermine (Fig. 1c) or 10% SP (Fig. 1d), and determined cell viability two days later. As expected, FCS alone had no effect on cell viability (Fig. 1c). Similarly, in the absence of FCS, spermine did not result in measurable cytotoxic activity (Fig. 1c). However, addition of only 1% FCS to cells exposed to spermine resulted in complete cell death (Fig. 1c). Similar results were obtained for 10% SP, which already caused 50% cell death in the absence of FCS, probably due to oxidases produced by the cells or naturally present in SP (Fig. 1d). These results suggest that spermine toxicity is dependent on the presence of FCS and that serum-free conditions may restore cell viability in the presence of otherwise toxic spermine concentrations.
We then analyzed the cytotoxicity of spermidine and spermine under serum-free cell culture conditions. For this, cells were either supplemented with a chemically-defined serum-free medium (“- serum” condition) or 10% FCS as control (“+ serum”), and then exposed to serial dilutions of spermidine and spermine. MTT tests performed 1, 2, and 3 days later confirmed strong cytotoxic effects of both polyamines at concentrations of ≥ 100 µM in the presence of 10% FCS (Fig. 2a, b). In contrast, under serum-free conditions, spermidine and spermine showed no strong cytotoxic effects at concentrations up to 5 mM, and reduced metabolic activity was only detected at the highest tested concentration of 10 mM (Fig. 2a, b). Thus, avoidance of FCS and utilization of serum-free medium allows studying spermine and spermidine at concentrations almost equivalent to those in semen.
To examine if amine oxidases in FCS are responsible for the conversion of spermine and spermidine into toxic intermediates, we analyzed the effect of the oxidase inhibitor AG. Control experiments showed that AG alone did not exert cytotoxic effects at concentrations of up to 5 mM in TZM-bl cells and primary blood mononuclear cells (PBMC) (Fig. S2), confirming previous data obtained with AG in other cells.17,24,26,34 We then incubated 100% FCS with 0, 50, 500 and 5,000 µM of AG for 24 hours and supplemented TZM-bl cells with 10% (v/v) of these samples together with spermidine and spermine concentrations of up to 10 mM. Cells were then incubated and MTT assays performed 2 days later. This experiment showed that AG concentrations of 50, 500 and 5,000 µM effectively reverted FCS-mediated cytotoxicity of both polyamines (Fig. 3a, b). These results confirm that amine oxidases in FCS are responsible for generating toxic polyamine products, as previously suggested.11,17−19 Furthermore, we confirm AG as supplement that enables analysis of high spermine and spermidine concentrations in the presence of FCS.
We were suspecting that the use of chemically-defined serum-free medium and/or AG-treated FCS may not only prevent cytotoxic effects of spermine and spermidine, but also SP. To test this, TZM-bl cells were either incubated with the standard supplement of 10% FCS (FCS), AG-treated FCS (FCS preincubated with 0.5 mM AG), chemically-defined medium (no FCS), or AG-treated chemically defined medium (no FCS, 0.05 mM AG). Cells were then exposed to SP concentrations of up 40% (v/v) and cell viability was determined 2 days later by MTT assay. As shown in Fig. 4a, in the presence of FCS, SP was strongly cytotoxic at concentrations of 0.3% and higher, as expected (Fig. 1a). When cells were supplemented with AG-treated FCS, SP was less toxic, with still more than 90% viable cells in the presence of up to 10% of SP (Fig. 4a). Serum-free conditions also allowed to analyze SP at concentrations of up to 2.5% (Fig. 4a). A combination of both, serum-free medium and AG, prevented toxic effects of SP most efficiently, allowing the analysis of SP concentrations of up to 10 volume% without causing any significant reduction in cell viability (Fig. 4a). Similar results were obtained using primary PBMCs instead of immortalized cells (Fig. 4b). In the presence of serum, SP caused massive cell death after 3 days, even at concentrations of only 0.2% SP (Fig. 4b). Utilization of AG-treated FCS largely prevented toxicity, allowing to study SP at concentrations of up to 10% in PBMCs. Again, a combination of serum-free medium and AG most effectively reduced SP-induced toxicity (Fig. 4b).
Finally, we evaluated whether addition of AG and utilization of serum-free medium may allow preventing SP toxicity in vaginal tissue blocks. For this, tissues derived from 4 donors were dissected into 2 x 2 x 1 mm3 blocks, and 10 blocks per donor were cultivated in either medium containing 10% FCS or serum-free medium containing 0.05 mM AG. Blocks were then exposed for three days to buffer or 10% and 20% SP (v/v) and cell viability was determined by measuring intracellular ATP levels using a luminescence-based assay (Fig. 5, Fig. S3), including a detergent control (0.5% triton). In the presence of FCS, 10% and 20% SP resulted in average in an 80% and 90% reduced viability as compared to the buffer control. Addition of AG completely rescued tissue viability in the presence of 10% SP and reduced viability by 25% in the presence of 20% SP. Thus, supplementation of medium with AG-treated FCS allows analysis of SP under conditions not affecting cell viability of vaginal tissue.