HIF1-alpha expressing cells induce a hypoxic-like response in neighbouring cancer cells

MCF7 and MDA-MB-231 were purchased from American Type Culture Collection (ATCC® HTB-22™ and HTB-26™). Cell lines were established for use in cell culture and no ethical approval was required for their use. Lines were authenticated by multiplex-PCR assay using the AmpF/STR system (Applied Biosystems) and confirmed as mycoplasma free. Monolayers were grown in complete medium (DMEM/10% FCS/2 mmol/L L-glutamine/PenStrep) and maintained in a humidified incubator at 37 °C at an atmospheric pressure of 5% (v/v) CO₂/air.

A yellow fluorescent protein tagged destabilising domain YFP-DD was removed from pBMN YFP-DHFR (DD) (a gift from Thomas Wandless, Addgene plasmid # 29326) and cloned into HA-HIF1-alpha P402A/P564A-pcDNA3 (a gift from William Kaelin, Addgene plasmid # 18955). The new vector, mHIF-YFP-DD-pcDNA3 was used to transfect MCF7 and MDA-MB-231 using Lipofectamine 2000 according to manufacturer’s instructions. Following selection with G418 cells were FACS sorted as single cells to produce stable, clonal cell lines. Proteins fused with YFP-DD are made unstable and although they are transcribed and translated continually, they are not seen to be expressed. Trimethoprim (TMP, Cay-16,473, Cayman) was used to induce stabilisation of the fusion.

Cells we incubated for 48 h in the SCI-tiveN hypoxic workstation (Ruskinn) in 1% O₂, 5% CO₂ and 94% N₂ in a humidified environment at 37 °C. Cells were plated, cultured, and harvested within the workstation to maintain hypoxia at all times. For cobalt chloride (CoCl₂) induction, 100 μM CoCl₂ was added to fresh culture medium and cells were incubated for 48 h in a humidified incubator at 37 °C at an atmospheric pressure of 5% (v/v) CO₂/air. Following hypoxic culture or exposure to CoCl₂, induction of HIF1-alpha was verified by Western blot (Additional file 1: Figure S4).

Conditioned medium (CM) was collected and spun at 1000 g for 10 min to remove cell debris. CM was stored at -20 °C for no longer than 2 weeks. CM was mixed 50:50 for fresh complete medium before being used to treat cells. The HIF1 inhibitor YC-1 was added to monolayer culture at a concentration of 10 mmol/L YC-1 (Cayman Chemicals).

Mammosphere culture was carried out as previously described [13], and spheres greater than 50 μm were counted on day 5. Briefly, cells were plated in non-adherent culture as single cell suspensions at a concentration of 2500 cells/6 well plate. Cells were cultured in DMEM:F12 supplemented with B27 (Thermo Fisher) for 5 days.

Parental and inducible cells were mixed at a ratio of 50:50 and cultured in the presence of TMP for 48 h. Cells were then separated based upon expression of YFP using the FACS Aria. When comparing to mono-culture all cells were passed through the sorter to account for any pressure/sorting effect.

Cells were plated on glass cover slides for 24 h. ClickIT™ EdU kit Plus EdU Alexa Fluor™ 647 Imaging Kit (C10640, Thermo) was used according to manufacturer’s instructions. Briefly cells were cultured with EdU for 45 min to allow incorporation of fluorescently labelled EdU into newly synthesized DNA therefore marking the proliferating cells. Cells were then fixed, permeablised and labelled with ClickIT-647.

Cells were resuspended in 400 μl of ice cold Buffer A (10 mM HEPES pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.5 mM PMSF) with the addition of complete mini-EDTA-free protease inhibitor cocktail (Roche) and incubated at 4 °C for 15 min. Cells were lysed with 10% NP-40 (Sigma) before centrifuging at 4 °C and removal of the cytoplasmic extracts in the supernatant. The pellet was then resuspended in ice cold Buffer B (20 mM HEPES pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT and 1 mM PMSF) containing protease inhibitors and vortexed vigorously for 45 min at 4 °C. Nuclear proteins were collected from supernatant following centrifugation at 4 °C. Both the nuclear and cytoplasmic extracts were stored at − 80 °C for future use.

Cells were grown in monolayer on coverslips for 48 h +/− 10 μM Trimethoprim (TMP). Medium was removed and cells were fixed and permeablised with 4% formaldehyde and 0.1% Triton for 20 min at room temperature. Non-specific binding was blocked using 1% BSA before antibody staining: rabbit anti-GFP (ab290, Abcam), Phalloidin-TRITC (P1951, SIGMA) or Phalloidin-iFluor 647 (ab176759, Abcam), Goat anti-Rabbit Alexa Fluor 488 (A11008, Thermo) and Goat anti-Mouse Cyanine3 (A10521, Thermo). Coverslips were mounted with ProLong™ Gold Antifade Mountant with DAPI (P36940, Thermo).

Images were collected on a Zeiss Axioimager.D2 upright microscope using a 10× objective and captured using a Coolsnap HQ2 camera (Photometrics) through Micromanager software v1.4.23. Specific band pass filter sets for DAPI, FITC and Cy5 were used to prevent bleed through from one channel to the next. Images were then processed and analysed using Fiji ImageJ (http://​imagej.​net/​Fiji/​Downloads) [14] which is freely available online.

Protein was separated on an SDS–PAGE and transferred to Hybond-C Extra nitrocellulose membrane. Primary antibodies included: HIF-1a (610,959, BD Biosciences) and Lamin B1 (ab16048, Abcam).

Densitometry was conducted using ImageJ software, which is freely available at http://​rsb.​info.​nih.​gov/​ij/​.

RNA was extracted using the Qiagen RNAeasy kit according to manufacturer’s instructions and quantified on the Nanodrop spectrophotometer (Thermo). Real time one step qRT-PCR was carried out using the QuantiTect SYBR® Green RT-PCR Kit (Qiagen) according to manufacturer’s instructions before analysis on the 7900 PCR machine (Applied Biosystems).

Data is represented as mean +/− SEM. Statistical significance was measured using parametric testing, assuming equal variance, in the majority of experiments with standard t tests for 2 paired samples used to assess difference between test and control samples.

To determine the effects of HIF1-alpha expressing cells on the surrounding HIF1-alpha negative cells we developed a model culture system in which HIF1-alpha can be induced in a sub-population of cells. To achieve this we fused a YFP-tagged destabilising domain (YFP-DD) to HIF1-alpha, thus enabling the stabilising effect of hypoxia on HIF1-alpha to be mimicked by the addition of an inducer molecule [15, 16]. The DD fusion protein is stable in the presence of Trimethoprim (TMP) but when TMP is removed the protein rapidly degrades (Fig. 1a). In this model YFP-DD was fused to a mutant form of HIF1-alpha (mHIF) which is stable in normoxia [17]. This fusion was stably expressed in MCF7 (ER+) and MDA-MB-231 (ER-) breast cancer cell lines [MCF7-mHIF-YFP-DD and 231-mHIF-YFP-DD] and inducible expression was verified by Western blot for HIF1-alpha performed on nuclear extracts following 48 h of culture with 10 μM TMP (Fig. 1b, Additional file 1: Figure S1 shows an unedited blot). Control lines were produced expressing YFP-DD only (MCF7-YFP-DD and 231-YFP-DD, data not shown).

To assess localisation and control of induction, MCF7-mHIF-YFP-DD and 231-mHIF-YFP-DD cells were cultured with 10 μM TMP for 48 h. Cells were then fixed and stained with an anti-GFP antibody to identify HIF1-alpha positive cells and Phalloidin-TRITC and DAPI to define the cytoplasm and nucleus respectively. Nuclear localisation, which is required for active HIF1-alpha signalling, was not affected by the increased size resulting from fusion to YFP-DD as the protein was present in the nucleus of induced cells (Fig. 1c). Activation of HIF1-alpha target genes was subsequently determined using qRT-PCR for three known genes. Significant increases in VEGF, EPO and CAIX were seen following the addition of 10 μM TMP (Fig. 1d). No change was observed in the control lines.

Induction occurred quickly with a significant increase in expression after 1 h in MCF7 cells and 30 min in MDA-MB-231. Expression reached its highest level by 24 h in both lines (Fig. 1e). Following removal and wash out of TMP, expression rapidly decreased returning to un-induced levels after 8 h. If TMP was not removed, induction remained high over 4 days following a single treatment with 10 μM TMP (Fig. 1f). Taken together these data show that the stabilising effect of hypoxia on HIF1-alpha can be mimicked by the addition of TMP in these cell lines.

We next determined if stabilising HIF1-alpha by the addition of TMP simulated the response known to be induced by hypoxia (1% O₂) and a chemical hypoxia mimetic (100 μM CoCl₂). qRT-PCR was performed on a panel of known HIF1-alpha target genes (VEGF, GLUT1, EPO, CAIX and TGFA) after 48 h induction with TMP, hypoxia or CoCl₂. In all three conditions target genes were induced to similar levels (Fig. 2a). TMP had no effect on gene expression (data not shown).

Hypoxic culture is known to have a contrasting effect on the cancer stem cell (CSC) population in breast cancer depending upon the disease sub-type [6]. In breast cancers which express estrogen receptor (ER), CSC activity increases following exposure to hypoxia. In contrast, there is a significant decrease in CSC activity in ER-negative tumours. This effect was previously shown to be directly down-stream of HIF1-alpha as siRNA targeting HIF1-alpha blocked the effect of hypoxic culture completely [6]. To assess whether the same effects were seen in our HIF1-alpha expressing cells, mammosphere formation was assessed following 48 h of culture with TMP. Mammosphere number was increased in MCF7-mHIF-YFP-DD cells and decreased in 231-mHIF-YFP-DD cells as is seen in hypoxic culture and following CoCl₂ exposure (Fig. 2b).

Secreted factors are known to play a major role in the regulation of hypoxia in tumour signalling both locally and to distant sites [18]. We aimed, therefore, to assess whether the CSC response seen following hypoxic culture or following HIF1-alpha induction in our model was controlled via secreted signals. Conditioned medium (CM) taken from cells cultured in hypoxia had the same effect on mammosphere formation as hypoxic culture itself regardless of the cell type being treated; CM taken from MCF7 (ER+) cells caused an increase in mammosphere number in both MCF7 and MDA-MB-231 whilst CM taken from MDA-MB-231 (ER-) caused a decrease in mammosphere formation in both MDA-MB-231 and MCF7 cells (Fig. 2c). This effect was blocked with addition of the HIF1-alpha inhibitor YC-1 during production of CM (Additional file 1: Figure S2). The same paracrine effects were seen when using CM collected from inducible lines; CM collected from MCF7-mHIF-YFP-DD and 231-mHIF-YFP-DD treated with TMP resulted in altered mammosphere formation in both MCF7 and MDA-MB-231 (Fig. 2c). Our data suggest that the HIF1-alpha induced mammosphere response is mediated by paracrine signals secreted by the hypoxic or HIF1-alpha expressing cells. The secreted factor (s) is different depending on the cell type releasing it and its effect is not cell type specific as CM from either cell line induced the same response in both parental lines. This supports the hypothesis that the hypoxic cancer cells within a tumour can influence the local micro-environment by affecting the different cell types surrounding them.

To simulate the heterogeneous nature of HIF1-alpha expression that is seen within a population of tumour cells we generated co-cultures by mixing wild-type MCF7 or MDA-MB-231 cells with HIF1-alpha inducible cells. TMP was added to co-cultures for 48 h before cells were isolated by flow sorting to separate HIF1-alpha positive (YFP positive) from negative cells (Additional file 1: Figure S3). These cultures were subsequently used to examine the effects of the HIF1-alpha positive cells on the surrounding parental sub-population.

Following co-culture and separation, MCF7 and MDA-MB-231 cells were plated into non-adherent culture to assess the effects on mammosphere formation. Induction of HIF1-alpha in mono-culture resulted in increased and decreased mammosphere formation as expected in MCF7 and MDA-MB-231 respectively (Fig. 3a). Importantly we observed that TMP induction of HIF1-alpha in MCF7 co-cultures also resulted in a significant increase in mammosphere formation in the non-induced sub-population of cells (Fig. 3a). In MDA-MB-231 co-cultures a reduction in mammospheres was observed. Moreover, qRT-PCR analysis of the HIF1-alpha target genes also revealed a significant increase in their expression in the parental, co-cultured cells (Fig. 3b). This included an increase in TGF-alpha expression in the parental cell lines following co-culture, a change which is not seen in the HIF1-alpha positive inducible cells grown in either mono- or co-culture or in those cultured in hypoxia (1%) or CoCl₂ culture. All genes responded similarly to HIF1-alpha induction in the HIF1-alpha expressing cells in both mono- and co-cultures (Fig. 3b). To determine the effects of HIF1-alpha induction on the proliferative capacity of neighbouring cells, ClickIT Edu was used to assess changes in proliferation after HIF1-alpha induction in co-cultured cells. HIF1-alpha negative cell proliferation was altered in MCF7 and MDA-MB-231 when grown in co-culture with HIF1-alpha expressing cells; MCF7 cells proliferated more quickly whilst MDA-MB-231 cells grew more slowly (Fig. 3c). HIF1-alpha has been shown to reduce proliferation in cancer cells [19] and, as expected, the HIF1-alpha positive cells proliferated more slowly than the parental cells in monoculture of both lines. Interestingly, the HIF1-alpha expressing cells proliferated more quickly when grown in co-culture and in the case of MDA-MB-231 there was no significant difference between the parental and inducible cells in this case. This change suggests a bi-directional effect of co-culture on proliferation suggesting a great deal can be learned from this system.

Taken together, these data reveal that induction of HIF1-alpha expression in a sub-population of cancer cells has a significant effect on gene expression, the mammosphere forming population and the proliferative capacity of the surrounding cancer cells. Thus our model system has revealed a complex interplay between HIF1-alpha positive and negative cancer cells that are likely to have a significant effect on the development of a tumour.