Adipose tissue-derived stem cells secrete CXCL5 cytokine with neurotrophic effects on cavernous nerve regeneration

Abstract

Introduction: Previously we reported that paracrine actions likely mediated the therapeutic effects of adipose tissue-derived stem cells (ADSCs) on a rat model of cavernous nerve (CN) injury.

Aim: To identify potential neurotrophic factors in ADSC's secretion, test the most promising one, and identify the molecular mechanism of its neurotrophic action.

Methods: Rat major pelvic ganglia (MPG) were cultured in conditioned media of ADSC and penile smooth muscle cells (PSMCs). Cytokine expression in these two media was probed with a cytokine antibody array. CXCL5 cytokine was quantified in these two media by enzyme-linked immunosorbent assay (ELISA). Activation of Janus Kinase/Signal Transducer and Activator of Transcription (JAK/STAT) by CXCL5 was tested in neuroblastoma cell lines BE(2)C and SH-SY5Y as well as in Schwann cell line RT4-D6P2T by Western blot. Involvement of CXCL5 and JAK/STAT in ADSC-conditioned medium's neurotrophic effects was confirmed with anti-CXCL5 antibody and JAK inhibitor AG490, respectively.

Main outcome measures: Neurotrophic effects of ADSC and PSMC-conditioned media were quantified by measuring neurite length in MPG cultures. Secretion of CXCL5 in these two media was quantified by ELISA. Activation of JAK/STAT by CXCL5 was quantified by densitometry on Western blots for STAT1 and STAT3 phosphorylation.

Results: MPG neurite length was significantly longer in ADSC than in PSMC-conditioned medium. CXCL5 was secreted eight times higher in ADSC than in PSMC-conditioned medium. Anti-CXCL5 antibody blocked the neurotrophic effects of ADSC-conditioned medium. CXCL5 activated JAK/STAT concentration-dependently from 0 to 50 ng/mL in RT4-D6P2T Schwann cells. At 50 ng/mL, CXCL5 activated JAK/STAT time-dependently, peaking at 45 minutes. AG490 blocked these activities as well as the neurotrophic effects of ADSC-conditioned medium.

Conclusions: CXCL5 was secreted by ADSC at a high level, promoted MPG neurite growth, and activated JAK/STAT in Schwann cells. CXCL5 may contribute to ADSC's therapeutic efficacy on CN injury-induced ED.

Figure 1

ADSC secretion promotes MPG neurite growth. Rat MPG dorsocaudal region was cultured in plain DMEM (A, Control Medium) or PSMC-conditioned DMEM (B), or in ADSC-conditioned DMEM (C) plus anti-CXCL5 antibody (D) or AG490 (E). For clarity, outgrowth of neurites was outlined by dashed lines. For quantification, the lengths of the 5 longest neurites in each of 18 specimens per treatment group were measured and their average shown on the Y-axis in Panel F. Alternatively quantification was done by measuring the neurite growth areas in each of 18 specimens per treatment group, the averages of which are shown on the Y-axis in Panel G.

Figure 2

ADSC secrete CXCL5 at a high level. (A) Cytokine expression in the conditioned media of ADSC and PSMC was detected by the RayBio Rat Cytokine Antibody Array, whose key is shown at the bottom. Cytokine LIX (CXCL5) is circled in the arrays and in the key. (B) CXCL5 secretion in the conditioned media of ADSC and PSMC was quantified by ELISA. Each bar represents the average of 3 independent experiments and each measurement was done in triplicates.

Figure 3

CXCL5 activates JAK/STAT in Schwann cells. (A-C) Neuroblastoma cell lines BE(2)C (A) and SH-SY5Y (B) and Schwann cell line RT4-D6P2T (C) were treated with CXCL5 at 0, 5, and 50 ng/ml for 45 min and then analyzed by western blot for the expression of STAT1, phosphorylated STAT1 (pSTAT1), STAT3, and phosphorylated STAT3 (pSTAT3). β-actin served as control. (D) The ratio (in percentile) of pSTAT1 versus STAT1 expression for each treatment was determined by dividing the densitometric values of these two protein bands obtained from the western blots. Each bar represents the average of 3 independent experiments. (E) Same as in (D) except that the proteins are pSTAT3 versus STAT3.

Figure 4

CXCL5 activates JAK/STAT time-dependently and with specificity. (A) Schwann cell line RT4-D6P2T was treated with CXCL5 at 50 ng/ml for the indicate time and then analyzed by western blot for the expression of STAT1, phosphorylated STAT1 (pSTAT1), STAT3, and phosphorylated STAT3 (pSTAT3). β-actin served as control. (B) Same as in (A) except that AG490 was added to a final concentration of 100 nM prior to the addition of CXCL5. (C) The ratio (in percentile) of pSTAT1 versus STAT1 expression for each treatment was determined by dividing the densitometric values of these two protein bands obtained from the western blots. Each bar represents the average of 3 independent experiments. (E) Same as in (C) except that the proteins are pSTAT3 versus STAT3.