C57BL/6J p53^−/− mice were generously provided by Dr Arnold Levine. C57BL/6J wt mice were purchased from Charles River Laboratories (Wilmington, MA, USA). Nude mice were purchased from Taconic Farms (Hudson, NY, USA). All animal procedures were approved by the Animal Care and Use Committee of RWJMS. MDA- MB231 cells were obtained from American Type Culture Collection (Manassas, VA, USA http://www.atcc.org); pooled human MSCs were obtained from Lonza (Walkersville, MD, USA, http://www.lonza.com) and used in early passage (below passage 8). MDA-MB231 cells were cultured in DMEM (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS) and 100 U/ml penicillin G and 100 μg/ml streptomycin at 37°C in 5% CO₂. Human MSCs were expanded in MesenCult media with hMSC stimulatory supplements (Lonza) and 10% FBS. Antibodies used in these studies included p53 (SC263) (Santa Cruz Biotechnology, Santa Cruz, CA, USA); α-tubulin and p21 (Sigma-Aldrich, St. Louis, MO, USA); GAPDH (Trevigen, Gaithersberg, MD, USA).

A Falcon cell culture insert system along with companion 24-well tissue culture plate was used for the chemotaxis assay as described previously (9). The polyethylene terepthalate membrane, pore size 8 μM, was selected to allow passage of mammalian cells. The insert was removed aseptically and placed in the notch of each well using forceps. MSCs (1–2×10⁴) were plated in 500 μl of α-MEM (Invitrogen) supplemented with 10% heat-inactivated FBS and 1% penicillin/streptomycin, placed in the insert (top chamber). The bottom chamber contained either conditioned medium from tumor cells or control medium (DMEM containing 2% FBS and penicillin/streptomycin). Migration assays were stopped after 16 h and cells remaining on the top of the membrane were removed with a wet cotton swab. MCSs that had migrated through the membrane were stained with crystal violet. Stained cells were counted under high power magnification (×40). For some experiments Nutlin-3 (Sigma-Aldrich, St. Louis, MO, USA) or recombinant CXCL12 (R&D Systems, Minneapolis, MN, USA) were added to both the upper and lower chambers to the indicated final concentrations.

Human MSCs (1.5×10⁵ cells) were plated in α-MEM (Invitrogen) with 10% FBS. After overnight incubation hMSCs were transfected with small interfering RNA (siRNA) specifically targeting p53 or CXCL12 or a scrambled sequence serving as a control (Thermo Scientific) using Lipofectamine 2000 (Invitrogen). Experiments were performed 2 days after transfection. Expression constructs containing short hairpin RNA (shRNA) sequences targeting p53 were obtained from Santa Cruz. hMSCs were infected using the manufacturer’s protocols and the knockdown of p53 was confirmed by western blotting. Cells were allowed to recover for 24 h prior to performing experiments.

To obtain tumor conditioned medium, a ratio of 7.5×10⁴ tumor cells/700 μl of DMEM containing 2% heat-inactivated FBS and 1% penicillin/streptomycin (Invitrogen) were incubated overnight. Conditioned medium was collected and spun down at 1,200 rpm to remove cellular debris. Supernatant was filtered through a 0.45-μm steriflip filter (Millipore, Billerica, MA, USA) prior to use in experiments.

Human MSCs were cultured to 80% confluence in a 150-mm tissue culture dish. Cells were then treated with MDA-MB231 conditioned medium or control medium with 25 μM Nutlin-3 or vehicle [dimethyl sulfoxide (Sigma-Aldrich)] for 6 and 24 h. Following this treatment, cells were scraped from the dish and re-suspended in 200 μl of radio-immunoprecipitation assay (RIPA) buffer plus 10 μg/ml aprotinin, 10 μg/ml leupeptin and 1 mM phenylmethylsulfonyl fluoride (PMSF). After clearing the cell lysates by centrifugation (16,000 g, 20 min at 4°C), the protein concentration was determined (Pierce, IL, USA). After boiling for 10 min, lysates (20 μg) were resolved by SDS-polyacrylamide gel electrophoresis, transferred onto PDVF membrane (Millipore) and visualized by immunoblotting with antibodies of interest.

Human MSCs were treated with Nutlin-3 as described above. Cells were then collected and RNA was extracted from MSCs using the RNeasy Mini kit (Qiagen Inc., Valencia, CA, USA) following standard procedures and quantified using the NanoDrop (Thermo Fisher Scientific, Rockford, IL, USA). Messenger RNA (mRNA) was used to generate complementary DNA (cDNA) which was amplified with a one-step RT-PCR kit (Applied Biosystems Inc., Foster City, CA, USA) using the MX4000 Multiplex Quantitative PCR System (Stratagene, Cedar Creek, TX, USA). CXCL12 cDNA was amplified by a CXCL12 taqman gene expression assay (Hs00930455, Applied Biosystems Inc., Foster City, CA, USA) using 100 ng of total RNA as starting material. 18s rRNA was amplified by the internal control 18s rRNA taqman assay (Applied Biosystems Inc.) using 1 ng of total RNA. Each RNA sample was assayed in quadruplicate and relative cDNA levels were determined after normalization to the internal 18s rRNA control.

MSCs were isolated from the bone marrow of C57BL/6J p53^−/− mice as previously described (9). Briefly, mice were euthanized using bottled CO₂ inhalation and the bilateral femurs were dissected out using sterile technique. The femurs were washed in phosphate-buffered saline containing 2% fetal bovine serum (FBS). Bone marrow cells were then obtained by flushing the femurs with PBS with 2% FBS. Cells were then filtered through a 70-μm nylon mesh and plated in α-MEM with 10% FBS and penicillin/streptomycin and cultured for seven days and then used in experiments. For use as a control, murine MSCs were transfected with wild-type murine p53. The p53 wild-type (wt) expression plasmid was the generous gift of Dr Arnold Levine. The p53-wt plasmid has been previously described and encodes murine p53 and G418 resistance with an SV40 origin of replication (26). Wild-type MSCs were labeled using carboxyfluorescein diacetate succinimidyl ester (CFSE) (Invitrogen) and p53 knockout MSCs were labeled using CellTracker CM-Dil (Invitrogen) according to the manufacturer’s instructions. The breast cancer cell line MDA-MB-231 (American Type Culture Collection) was used in this study. Cells (106) along with Matrigel (50 μl per injection; BD Biosciences) were injected subcutaneously into nude mice and tumors were allowed to reach a size of ∼5 mm in diameter. At this point, CFSE-labeled murine MSCs expressing human p53 and CM-Dil-labeled murine p53^−/− MSCs were mixed at a ratio of 1:1. Mixed MSCs (5×10⁶) were then injected subcutaneously 10 mm from each tumor. Seven days later the mice were sacrificed using CO₂ inhalation. Tumors were dissected from the mice and a single cell suspension of each tumor was made. Tumors were dissected into small pieces using a scalpel and dissociated using collagenase (Roche, Mannheim, Germany) (0.035% wt/vol) in α-MEM at 37°C for 1 h. Tissue was further dissociated by pipetting several times through a 5cc pipette and debris was removed using a nylon strainer. The single cell suspension was them washed with α-MEM with 10% FBS. The cellular component of each tumor was then analyzed using flow cytometry.

At least three independent experiments were performed for each in vitro migration assay. Results are presented as the means ± standard deviation. Statistical significance was determined using the Student’s t-test and a value of P<0.05 was considered statistically significant. Microsoft Excel software was use for statistical analysis.

Human MSCs were treated with the murine double minute 2 (MDM2) antagonist, Nutlin-3, leading to expected increases in p53 as well as increases in the p53 target p21 (Fig. 1A). In conjunction with increased p53 levels, the in vitro migration of MSCs in response to MDA-MB-231 tumor cells was decreased. The migration to tumor conditioned media showed a non-significant trend toward decreasing and Nutlin-3 did not inhibit the migration of MSCs in response to interleukin 8 (IL-8) (Fig. 1B). There was minimal basal migration of MSCs in response to control medium and this was not changed by treatment with Nutlin-3. When levels of p53 were decreased using siRNA (Fig. 1C), MSCs exhibited increased migration in response to tumor cells (Fig. 1D). These results suggested that p53 plays a role in regulating the response of MSCs to tumor cells.

hMSCs were exposed to a combination of Nutlin-3 and MDA-MB-231 conditioned medium. As expected, exposure of MSCs to nutlin-3 led to increased levels of p53 as well as its target p21. However, exposure of MSCs to tumor conditioned medium did not impact p53 levels (Fig. 2). This result indicates that, while the p53 level influences MSC migration in response to tumor cells, induction of MSC motility by TCM is not mediated by changes in p53 activity.

We next explored the mechanism of regulation of MSC migration by p53. Because production of CXCL12 by MSCs is required for their migration in response to tumor cells (9), we investigated the effect of increased p53 on CXCL12 production by MSCs. Nutlin-3 treatment decreased hMSC CXCL12 mRNA levels after 24 h. CXCL12 mRNA levels in MSCs exposed to conditioned medium from MDA-MB-231 cells were decreased by nutlin-3 treatment at both 6- and 24-h time intervals (Fig. 3). These data suggested that p53 impacts MSC migration through regulation of CXCL12 transcription.

To further demonstrate the mechanism of action for Nutlin-3 inhibition of MSC chemokinesis, MSCs with p53-knockdown were treated with Nutlin-3. Nutlin-3 did not diminish the migration of MSCs with p53 knockdown induced by tumor cells (Fig. 4), indicating that p53 is required for Nutlin-3-mediated suppression of migration.

In order to determine whether the regulation of CXCL12 plays a role in the increased motility of MSCs with p53 knockdown, MSCs with both p53 and CXCL12 knockdown were generated using siRNA. CXCL12 knockdown blocked the increased motility of MSCs with p53 knockdown (Fig. 4).

We then sought to determine whether the effect of p53 levels on MSC migration was exclusively through its role in the regulation of CXCL12 production. Recombinant CXCL12 protein was added to human MSCs that had been treated with Nutlin-3. While exogenous application of CXCL12 does stimulate migration of MSCs in response to tumor conditioned medium (9), recombinant CXCL12 failed to increase migration of MSCs treated with Nutlin-3 (Fig. 5). This result suggests that, in addition to regulating the production of CXCL12, p53 may impact MSC migration through additional mechanisms.

To determine whether p53 regulates homing of MSCs to tumor sites in vivo, we used bone marrow-derived MSCs isolated from p53-null mice. The human wild-type p53 gene was introduced into the MSCs using a lentiviral vector. Western blot analysis demonstrated that MSCs isolated from p53-null animals were deficient in p53 protein and that after transfection of wild-type p53 gene, the expression of the tumor suppressor was detected (Fig. 6A). Wild-type MSCs were labeled using CFSE and p53 knockout MSCs were labeled using CM-Dil. Wild-type and p53^−/− MSCs were mixed in a 1:1 ratio and subcutaneously co-injected into nude mice 5 mm from established tumors (MDA-MB231). At day 7 after administration of the MSCs, animals were sacrificed and tumors were collected and used to generate single cell suspensions. The ratio of wild-type to p53^−/− cells in the tumor was then determined using flow cytometry. Increased number of p53-null MSCs were found in the tumors compared to wild-type MSCs, indicating that the in vivo homing capability of MSCs was enhanced in cells with decreased p53 activity (Fig. 6B).