Kallikrein 6 is a novel molecular trigger of reactive astrogliosis

Abstract

Kallikrein-related peptidase 6 (KLK6) is a trypsin-like serine protease upregulated at sites of central nervous system (CNS) injury, including de novo expression by reactive astrocytes, yet its physiological actions are largely undefined. Taken with recent evidence that KLK6 activates G-protein-coupled protease-activated receptors (PARs), we hypothesized that injury-induced elevations in KLK6 contribute to the development of astrogliosis and that this occurs in a PAR-dependent fashion. Using primary murine astrocytes and the Neu7 astrocyte cell line, we show that KLK6 causes astrocytes to transform from an epitheliod to a stellate morphology and to secrete interleukin 6 (IL-6). By contrast, KLK6 reduced expression of glial fibrillary acidic protein (GFAP). The stellation-promoting activities of KLK6 were shown to be dependent on activation of the thrombin receptor, PAR1, as a PAR1-specific inhibitor, SCH79797, blocked KLK6-induced morphological changes. The ability of KLK6 to promote astrocyte stellation was also shown to be linked to activation of protein kinase C (PKC). These studies indicate that KLK6 is positioned to serve as a molecular trigger of select physiological processes involved in the development of astrogliosis and that this is likely to occur at least in part by activation of the G-protein-coupled receptor, PAR1.

Figure 1. KLK6 is dense in hypertrophic astrocytes in spinal cord injury, multiple sclerosis and glioblastoma multiforme

In normal appearing spinal cord white matter (A, B), KLK6 is prominent in oligodendrocytes which bear round nuclei and a thin rim of cytoplasm (arrows, A), while astrocytes are not prominently stained(Scarisbrick et al., 2000; Scarisbrick et al., 2001) (B, shows adjacent LFB/PAS). In cases of traumatic spinal cord injury (C) and in multiple sclerosis lesions (E), KLK6 immunoreactivity is dense in hypertrophic astrocytes (D and F show Luxol fast blue/periodic acid Schiff stains in near adjacent sections demonstrating myelin loss; small arrows A and C indicate KLK6-immunoreactive oligodendrocytes, large arrows C and E indicate hypertrophic astrocytes). KLK6 is also dense in astrocytes with a gemistocytic morphology seen in glioblastoma multiforme tumor samples (G, large arrows) (Scale Bar = 50 μm in A–G).

Figure 2. KLK6 promotes stellation of primary murine cortical astrocytes

Primary astrocytes were treated with (A) vehicle alone, (B) KLK6 (300 nM (10 μg/ml)), (C) KLK1 (300 nM (10 μg/ml)), (D) PAR1-AP (40 μM), (E) PAR2-AP (200 μM); or (F) LPS (25 μg/ml) for 24 hr and stained with rhodamine-conjugated Phalloidin to visualize actin. (G) Histogram shows percent stellation induced by each agonist. While KLK6-triggered robust stellation, neither KLK1, PAR1- or PAR2-APs increased stellation above baseline levels. LPS was used as a positive control for astrocyte activation and promoted stellation. (H) KLK6-induced stellation (300 nM) was observed as early as 4 hr after KLK6 treatment. (I) Significant KLK6-mediated stellation effects were seen at both 30 and 300 nM (1 and 10 μg/ml). (J) Aprotinin (Apro) blocked KLK6-induced astrocyte stellation (Asterisks p<0.05, SNK). (Scale Bar = 50 μm in A–F)

Figure 3. KLK6 promotes stellation of Neu7 astrocytes

The Neu7 astrocyte cell line was treated with (A) vehicle alone, (B) KLK6 or (C) KLK1 each at 300 nM (10 μg/ml), (D) PAR1-AP (40 μM), (E) PAR2-AP (200 μM), or (F) PAR1- in addition to PAR2-APs and stained with rhodamine-conjugated Phalloidin to visualize actin. (G) KLK6 significantly increased stellation above baseline, while treatment with PAR1-, PAR2-, or combined PAR1- and PAR2-APs, caused significant stellation reversal. KLK1 did not significantly alter stellation. (H) KLK6-triggered stellation was associated with a significant increase in astrocyte process length. (I) Less PAR1-AP mediated stellation reversal was observed when cells were concomitantly treated with KLK6. The ability of KLK6 to decrease PAR1-mediated stellation reversal was blocked by the serine protease inhibitor APMSF. (Asterisks p<0.05, SNK). (Scale Bar = 100 μm in A-F).

Figure 4. KLK6-induced astrocyte stellation is dependent on PAR1

Histograms show the PAR1 inhibitor SCH79797 (35 nM) blocks astrocyte stellation induced by (A) 30 nM (1 μg/ml) KLK6 in primary astrocytes and (B) 330 nM (10 μg/ml) KLK6 in the Neu7 astrocyte cell line. PAR1-inhibitor mediated reductions in KLK6-stellation seen at higher concentrations of KLK6 in primary astrocytes did not reach the level of statistical significance. (Asterisks p<0.05, SNK).

Figure 5. KLK6-induced astrocyte stellation is blocked by inhibition of PKC signaling

Histograms show that the PKC inhibitor Go6983 (60 nM) reduces the ability of KLK6 (300 nM (10 μg/ml)) to induce stellation in (A) primary astrocytes or (B) the Neu7 astrocyte cell line. (Asterisks p<0.05, SNK).

Figure 6. KLK6 promotes secretion of IL-6 but reduces expression of GFAP mRNA in primary murine astrocytes

(A) Treatment of primary astrocytes for 24 hr with KLK6 (300 nM (10 μg/ml)) resulted in a significant increase in IL-6 secretion but a decrease in expression of GFAP mRNA (B). (C) In primary astrocyte experiments (see Figure 2), counts of the mean number of DAPI labeled nuclei reveal no significant impact of KLK6-treatment (300 nM) on cell number at the 24 hr endpoint of each experiment. (Asterisks p<0.05, SNK).