The IκB kinase inhibitor ACHP strongly attenuates TGFβ1‐induced myofibroblast formation and collagen synthesis

Abstract Excessive accumulation of a collagen‐rich extracellular matrix (ECM) by myofibroblasts is a characteristic feature of fibrosis, a pathological state leading to serious organ dysfunction. Transforming growth factor beta1 (TGFβ1) is a strong inducer of myofibroblast formation and subsequent collagen production. Currently, there are no remedies for the treatment of fibrosis. Activation of the nuclear factor kappa B (NF‐κB) pathway by phosphorylating IκB with the enzyme IκB kinase (IKK) plays a major role in the induction of fibrosis. ACHP {2‐Amino‐6‐[2‐(cyclopropylmethoxy)‐6‐hydroxyphenyl]‐4‐(4‐piperidinyl)‐3 pyridinecarbonitrile}, a selective inhibitor of IKK, prohibits the activation of the NF‐κB pathway. It is not known whether ACHP has potential anti‐fibrotic properties. Using adult human dermal and lung fibroblasts we have investigated whether ACHP has the ability to inhibit the TGFβ1‐induced transition of fibroblasts into myofibroblasts and its excessive synthesis of ECM. The presence of ACHP strongly suppressed the induction of the myofibroblast markers alpha‐smooth muscle actin (αSMA) and SM22α, as well as the deposition of the ECM components collagen type I and fibronectin. Furthermore, post‐treatment with ACHP partly reversed the expression of αSMA and collagen type I production. Finally, ACHP suppressed the expression of the three collagen‐modifying enzymes lysyl hydroxylase (PLOD1,PLOD2 and PLOD3) in dermal fibroblasts, but did not do so in lung fibroblasts. We conclude that the IKK inhibitor ACHP has potent antifibrotic properties, and that the NF‐κB pathway plays an important role in myofibroblast biology.


Introduction
Fibrosis is a common outcome of an impaired tissue repair process. The hallmark of fibrosis is the production and excessive accumulation of a collagen-rich extracellular matrix (ECM). The ECM is deposited by myofibroblasts under the influence of pro-fibrotic cytokines, such as transforming growth factor beta 1 (TGFb1) [1,2]. TGFb1 is one of the most potent pro-fibrotic cytokines known to be involved in the activation of fibroblasts into myofibroblasts, a key process associated with fibrosis that seems to be dependent, among others, on the activation of the nuclear factor kappa B (NF-jB) pathway [3][4][5][6], especially NF-jB subunit p65 (= RelA).
The RelA signalling pathway is associated with the pathogenesis of fibrosis including organs such as kidney, liver and lung [5][6][7][8][9][10][11][12][13]. In unstimulated cells, the cytoplasmic NF-jB is bound to the inhibitory protein IjB; the complex prevents NF-jB activation (i.e. the complex remains in the cytoplasm). Extracellular stimuli activate the enzyme IjB kinase (IKK) that phosphorylates the IjB subunit of the NF-jB/ IjB complex. The phosphorylated IjB is degraded and NF-jB is translocated to the nucleus, where it is able to bind to its target sequences and subsequently activate gene transcription [13][14][15][16].
Activation of IKK has been reported to induce liver fibrosis in mice [12]. Selective deletion of IKKb (a catalytic subunit of the IKK complex) in mouse airway epithelium blocked the nuclear translocation of RelA and showed less peribronchial fibrosis [17]. Overexpression of integrin-linked kinase in rat cardiac fibroblasts induces a pro-fibrotic response by increasing the production of collagen type I and connective tissue growth factor via RelA; transfection of cardiac fibroblasts with mutant IjBa (a member of IjB) inactivated RelA, thereby decreasing fibrosis [18]. A direct approach using RelA antisense oligonucleotides reduced the formation of the myofibroblast marker alpha-smooth muscle actin (aSMA) in bleomycin-induced mouse lung fibrosis and in cultured cells, showing the deleterious role of NF-jB in the development and progression of organ fibrosis [9,19].
These examples show that inhibition of the IKK/NF-jB pathway might be an attractive therapeutic tool to attenuate fibrosis.
Several NF-jB pathway inhibitors have been investigated in animal models to slow down the fibrotic reaction. IMD-0354 (an IKKb inhibitor) prevented the activation of RelA and collagen content in bleomycin-induced lung fibrosis in mice [20]. Administration of Suramin, a polysulfonated naphthylurea, inhibited the TGFb1/ Smad3 pathway and the phosphorylation of IjBa in fibrotic peritoneum and thereby reduced peritoneal fibrosis in rat [21]. Salvianolic acid B, derived from Salvia miltiorrhiza (a Chinese herbal medicine), has been reported to reduce carbon tetrachloride-induced liver fibrosis in rats which correlated with an increased level of RelA and IjBa protein in the cytoplasm but not in the nucleus [22]. Similarly, the expression of aSMA was decreased through the inhibition of IKK with a boswellic acid-containing extract treatment in a mouse model of Schistosomiasis liver granuloma and fibrosis [23]. Pressure overload-induced cardiac fibrosis has been treated with Sophocarpine, a tetracyclic quinolizidine alkaloid, resulting in a reduction of collagen deposition by inhibiting IjBa phosphorylation [24]. Unfortunately, in all these studies, only limited protective effects of these agents have been described. Therefore, another therapeutic agent that interferes with the NF-jB system is wanted to reduce or inhibit fibrosis progression.
The low molecular weight compound 2-Amino-6-[2-(cyclopropylmethoxy)-6-hydroxyphenyl]-4-(4-piperidinyl)-3 pyridinecarbonitrile (ACHP) is a selective inhibitor of IKK (both for the IKKa and the IKKb subunit) [25][26][27]. So far, no investigations have been performed to explore whether ACHP is able to interfere with fibrotic processes, such as blocking the TGFb1-induced transition of fibroblasts into myofibroblasts. In this study, we examined whether ACHP can directly inhibit myofibroblast formation and ECM synthesis. To decipher this, adult human dermal and lung fibroblasts (HDFa and HLFa) were stimulated with TGFb1 in the presence or absence of ACHP and investigated the formation of myofibroblasts and the deposition of ECM-molecules. In addition, we explored whether myofibroblasts that are formed by TGFb1 can be reversed into fibroblasts with an ACHP post-treatment. We found that ACHP strongly attenuates TGFb1-induced formation of myofibroblasts as well as collagen type I and fibronectin protein synthesis.

Cell culture
Human adult dermal fibroblasts [Caucasian, 20 years, CCD-1093Sk (ATCC â CRL-2115 TM ), ATCC, Manassas, VA, USA] and Human adult lung fibroblasts [Caucasian, 27 years, CCD-19Lu (ATCC â CCL-210 TM ), ATCC] were cultured in basal medium (= EMEM containing 1% L-glutamine and 1% penicillin/streptomycin) supplemented with 10% FBS. Passage 5 to 8 of HDFa and HLFa were seeded with a density of 15,000 cells/cm 2 in a Costar 12-well plate for quantitative real-time polymerase chain reaction (qRT-PCR) or in a 48-well plate for immunofluorescence staining. After 72 hrs, fibroblasts were washed with PBS and starved overnight with basal medium containing 0.5% FBS. Subsequently, fibroblasts were trea-  ted with/without ACHP (50 lM), recombinant human TGFb1 (10 ng/ml), or a combination of both, for a period of 24 hrs (for qRT-PCR) and 48 hrs (for immunofluorescence staining) in basal medium supplemented with 0.17 mM L-ascorbic acid 2-phosphate magnesium salt and 0.5% FBS. In another experiment, fibroblasts were stimulated with TGFb1 (10 ng/ml) for 48 hrs followed by a post-treatment with/without ACHP (50 lM) for 24 hrs in basal medium supplemented with 0.17 mM Lascorbic acid 2-phosphate magnesium salt and 0.5% FBS. Subsequently, whole-cell lysates (as obtained with FARB-buffer) were used for qRT-PCR. For immunofluorescence analysis, fibroblasts were washed with PBS and fixed with methanol/acetone solution (1:1 ratio) for 5 min. The ACHP compound was dissolved in sterile water at a A C D E B Fig. 2 Effects of ACHP on transforming growth factor beta1 (TGFb1)-induced alpha-smooth muscle actin (aSMA) synthesis. (A and B) human dermal and lung fibroblasts (HDFa and HLFa) were cultured for 24 hrs in the presence of ACHP alone, TGFb1 alone or TGFb1 in combination with ACHP (co-treatment). mRNA levels of ACTA2 relative to the reference gene YWHAZ. (C-E) HDFa and HLFa were cultured for 48 hrs in the presence of ACHP alone, TGFb1 alone or TGFb1 in combination with ACHP (co-treatment). Representative immunofluorescence stainings (left panel) and quantification of the % of cells (right panel) positive for aSMA. The scale bars represent 100 lm. The sign * represents statistically significance towards untreated control, and the sign # represents statistically significance of cells co-stimulated with TGFb1 and ACHP compared to TGFb1-treated cells. Fig. 1 Effects of ACHP on the nuclear translocation of RelA and Smad2/3. (A and B) human dermal and lung fibroblasts (HDFa and HLFa) were cultured for 30 min. in the presence of ACHP alone, transforming growth factor beta1 (TGFb1) alone or TGFb1 in combination with ACHP (co-treatment). Representative immunofluorescence stainings for the cytoplasmic and nuclear localization of RelA with confocal laser scanning microscopy. (C and D) HDFa and HLFa were cultured for 45 min. in the presence of ACHP alone, TGFb1 alone or TGFb1 in combination with ACHP (co-treatment). Representative immunofluorescence stainings for the cytoplasmic and nuclear localization of Smad2/3 with confocal laser scanning microscopy. The scale bar represent 25 lm. RNA isolation, cDNA synthesis and qRT-PCR Total RNA was isolated using the Favorgen RNA extraction kit and reverse transcribed with the First Strand cDNA synthesis kit. Gene expression analysis was performed by means of qRT-PCR in a 10 ll reaction mixture containing 10 ng cDNA, SYBR Green Master Mix, 6 lM forward primer and 6 lM reverse primer (for primer sequences see Table 1). qRT-PCR was conducted in triplicate for each condition in a 384-well plate at 95°C for 15 sec. and 60°C for 1 min. for 40 cycles using the ViiA 7 Real-Time PCR System (Applied Biosystems, Waltham, MA, USA). All mRNA data were normalized against the reference gene tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, zeta isoform (YWHAZ).

Statistics
All mRNA and immunofluorescence data are presented as mean AE SEM for at least three independent experiments. Results were analysed with either one-way ANOVA followed by Tukey's post-test or two tailed unpaired t-test analysis using Graph-Pad Prism Version 5 (GraphPad Software, La Jolla, CA, USA). P < 0.05 was considered to be statistically significant. The signs * and # represents a statistically significant difference compared to the untreated, and TGFb1 treatment respectively.

ACHP inhibits nuclear translocation of RelA
First, we examined whether ACHP indeed blocked the enzyme IKK. If so, RelA in HDFa and HLFa would mainly stay in the cytoplasm, instead of being translocated to the nucleus. Stimulation of HDFa and HLFa with TGFb1 induces the nuclear translocation of RelA first at 15 min. of treatment (data not shown), and reached the highest nuclear expression after 30 min. of treatment ( Fig. 1A and B). Treatment of fibroblasts with ACHP in presence of TGFb1 efficiently inhibited the nuclear translocation of RelA in a time-dependent manner. The inhibitory effects were seen as early as 15 min. after treatment (data not shown) and the effect was most prominent at 30 min. (Fig. 1A and B).

Effects of ACHP on the activation of Smad2/3
Since RelA is involved in the myofibroblast differentiation process as induced by TGFb [5], we then wondered whether a downstream target of TGFb1-signalling, namely Smad2/3, is affected by the lack of nuclear translocation of RelA. We thus evaluated the effects of ACHP on nuclear translocation of Smad2/3 in HDFa and HLFa. Confocal microscopy imaging shows that TGFb1-treatment activated the nuclear translocation of Smad2/3 in both types of fibroblasts already at 15 min. of incubation (data not shown) and reached the highest level at 45 min. (Fig. 1C and D). We therefore selected 45 min. to observe the effects of ACHP on Smad2/3 translocation. In a TGFb1-rich pro-fibrotic environment, ACHP resulted in a blockage of the nuclear translocation of Smad2/3 in both HDFa and HLFa (Fig. 1C and D).

Effect of TGFb1 and ACHP on aSMA synthesis
Myofibroblasts are the crucial cell types involved in ECM production and tissue contraction. These cells are enriched with cytoplasmic stress fibres containing aSMA and smooth muscle protein 22alpha (SM22a) [1][2][3][4]28]. To examine whether ACHP can inhibit the TGFb1-induced differentiation of fibroblasts into myofibroblasts, HDFa and HLFa were incubated either with ACHP, TGFb1 or both. Gene expression analysis of TGFb1-stimulated fibroblasts showed an up-regulation (~7-fold) in mRNA level of ACTA2 (the gene that encodes for the protein aSMA) compared to unstimulated fibroblasts ( Fig. 2A and B). Similarly, the immunofluorescence staining on aSMA protein showed an increase in the number of aSMA-positive cells (in HDFa~12% and in HLFa~70%) after stimulation with TGFb1. Interestingly, the combination of ACHP and TGFb1 resulted in a decrease in aSMA synthesis to baseline levels, both on a gene and on a protein level. Treatment of ACHP alone also reduced the mRNA and protein level of aSMA in unstimulated HLFa, whereas ACHP had no effect on HDFa (since there are hardly aSMA-positive cells detected in the pool of unstimulated HDFa) ( Fig. 2A-E). These results reveal that ACHP is able to inhibit both the pre-existing and the TGFb1-induced synthesis of aSMA stress fibres. In addition, to evaluate whether ACHP is able to reverse TGFb1-induced formation of aSMA, HDFa and HLFa were pre-treated with TGFb1 followed by ACHP incubation. TGFb1-treatment alone resulted in an increase in positively stained aSMA fibroblasts (in HDFa~25% and in HLFa~65%) and the post-treatment with ACHP showed a~50% reduction in aSMA-positive cells in HDFa, whereas HLFa did not show such a decreasing effect on aSMA stress fibres (Fig. 3).

Effect of TGFb1 and ACHP on SM22a synthesis
Myofibroblasts are also characterized by increased levels of SM22a, another component of cytoplasmic stress fibres [28]. Like aSMA, we investigated whether ACHP is able to prevent the synthesis of SM22a in a TGFb1-enriched environment. Compared to unstimulated fibroblasts, stimulation of HDFa and HLFa with TGFb1 showed an up-regulation of TAGLN mRNA (in HDFa~4-fold and in HLFa~2-fold), the gene that encodes for the protein SM22a. As was the case for ACTA2, this up-regulation was blocked by ACHP to baseline levels both in HDFa and HLFa (Fig. 4A and B). The protein expression of SM22a showed a similar result as seen with mRNA expression. In comparison to untreated fibroblasts, SM22a stress fibres were up-regulated (~2-fold) with TGFb1 for both fibroblast types; co-treatment with ACHP suppressed the number of SM22a-positive cells to baseline levels, both for HDFa and HLFa. Incubation of ACHP alone revealed a downregulation of SM22a synthesis both at the gene and protein level in HLFa, whereas in HDFa a reduction was seen only at mRNA level ( Fig. 4C-E). These data show, together with the above data on aSMA, that ACHP is able to block the TGFb1-induced formation of myofibroblasts.

Effect of TGFb1 and ACHP on collagen type I synthesis
Increased production and deposition of collagen type I by myofibroblasts is the main feature of fibrosis [29,30]. We investigated whether ACHP was able to block the synthesis of collagen type I. In TGFb1stimulated fibroblasts an increase in mRNA level of COL1A1 (the gene that encodes the a1 chain of collagen type I) is seen in either type of fibroblasts (~2.5-fold in HDFa and~1.5-fold in HLFa) compared to untreated controls. Combination treatment of ACHP with TGFb1 suppressed the expression of COL1A1 to baseline levels both for HDFa and HLFa (Fig. 5A and B). A similar outcome was observed regarding collagen type I protein synthesis. Compared to untreated fibroblasts, a~50% increase in the number of collagen-producing cells was observed after TGFb1-stimulation, whereas co-stimulation with ACHP blocked the synthesis of collagen type I to baseline for both fibroblast types (Fig. 5C-E). ACHP alone also inhibited mRNA expression of COL1A1 in unstimulated HDFa and HLFa ( Fig. 5A and B). Protein expression of collagen type I was completely abolished by ACHP in HLFa, but such an effect could not be observed in HDFa as there are hardly any cells positive for collagen type I expression in the unstimulated pool (Fig. 5C-E). Post-treatment with ACHP after TGFb1 treatment resulted in a major decrease (~30% in HDFa and~25% in HLFa) of the number collagen-expressing fibroblasts (Fig. 6). Taken together, these findings indicate that ACHP can at least slowdown the accumulation of collagen type I.

Effect of TGFb1 and ACHP on fibronectin synthesis
The adhesion protein fibronectin (FN1) is an important ECM molecule required for the attachment and migration of fibroblasts and facilitates the construction of an organized matrix in tissue reparative processes. Aberrant expression of FN1 under the influence of TGFb1 contributes to fibrosis [2]. Therefore, we have examined whether ACHP is able to prevent the TGFb1-induced production of fibronectin. Incubation of fibroblasts with TGFb1 showed a~3-fold and~2-fold increase in the mRNA level of FN1 in HDFa and HLFa, respectively. Co-treatment with ACHP inhibited this up-regulation, reaching baseline levels for both HDFa and HLFa ( Fig. 7A and B). Immunofluorescent staining exhibited a 2-fold increase in the number of fibronectin-positive fibroblasts with TGFb1 compared to untreated cells. TGFb1 in combination with ACHP notably reduced the number of fibronectin-positive cells, both for HDFa and HLFa (Fig. 7C-E). Treatment of fibroblasts with ACHP alone slightly inhibited the gene expression of FN1; this was not reflected in protein expression level for either type of fibroblasts ( Fig. 7A-E). These results show that ACHP has an inhibitory effect on the synthesis of fibronectin.

Effect of TGFb1 and ACHP on the synthesis of members of the PLOD family
The processing and modification of collagen by lysyl hydroxylase (LH) enzymes plays a critical role in the stability of the collagen fibril.
In fibrosis, a highly elevated level of the collagen-modifying enzyme LH2 encoded by the gene PLOD2 has been reported [31][32][33][34], whereas up-regulation of LH1 (encoded by PLOD1) and LH3 (encoded by PLOD3) is more modest. We wondered whether ACHP has an effect on the expression of the PLOD family members in our experimental setting. Compared to unstimulated cells, TGFb1-stimulation significantly increased mRNA levels of PLOD1 and PLOD2 in both fibroblast types; the fold-increase for PLOD2 was always higher than for PLOD1. Little or no increase in PLOD3 expression was seen in HDFa and HLFa, respectively. The presence of ACHP during TGFb1 incubation showed a downregulation of all three PLODs in HDFa, whereas no down-regulation was seen in HLFa. Similar observations were made when unstimulated fibroblasts were treated with ACHP alone (Fig. 8A-F).

Effects of ACHP on proliferation of myo (fibroblasts)
Lastly, we investigated the effects of ACHP on the proliferation of myo(fibroblasts). Staining for the proliferation marker Ki-67 showed that there was a significant increase in the proliferation of HDFa and HLFa upon TGFb1-stimulation compared to untreated fibroblasts. ACHP markedly reduced the number of TGFb1-induced Ki-67 positive cells in both HDFa and HLFa. However, no difference was observed between the proliferation rates of untreated cells and cells treated with ACHP alone (Fig. 9A and B), indicating that the given concentration of ACHP is not toxic for the cells (as can also be concluded from the normal morphological appearance of the ACHP-treated cells). Fig. 9 Effects of ACHP on the proliferation of myo(fibroblasts). (A and B) Human dermal and lung fibroblasts (HDFa and HLFa) were cultured for 48 hrs in the presence of ACHP alone, transforming growth factor beta1 (TGFb1) alone or TGFb1 in combination with ACHP (co-treatment). Quantification of the % of cells positive for Ki-67. The sign * represents statistically significance towards untreated control, and the sign # represents statistically significance of cells co-stimulated with TGFb1 and ACHP compared to TGFb1-treated cells.

Discussion
The strong pro-fibrotic effects of TGFb1 towards fibroblasts results in the formation of myofibroblasts, being the critical cell type that plays a central role in fibrosis. Myofibroblasts are responsible for the synthesis of large amounts of ECM [1][2][3][4]28]. Here, we show that ACHP is able to completely abrogate the TGFb1-induced differentiation of both dermal and lung fibroblasts into myofibroblasts, as revealed by the inhibition of the formation of the stress fibre components aSMA and SM22a. The data on the post-treatment by ACHP revealed a 50% decrease in TGFb1-induced aSMA-expressing dermal fibroblasts, whereas no decrease was observed in lung cells. These findings indicate that ACHP is able to inhibit myofibroblast formation of HDFa and HLFa in a TGFb1-rich environment. In the case of HDFa, ACHP was even able to partially reverse myofibroblast into fibroblasts. The diminished nuclear translocation of NFjB, as induced by ACHP, has apparently major implications on the TGFb pathway itself. Indeed, we showed also a lack of nuclear translocation of Smad2/3, a downstream target of TGFb, thus providing further mechanistic evidence on how NFjB regulates myofibroblast differentiation [5,9,19].
Excessive collagen production is the hallmark of fibrosis [29,30]. We have found that ACHP blocks the synthesis of TGFb1-induced collagen type I to baseline values, both at gene and protein level, and independent of fibroblast type. A considerable reduction in synthesized collagen type I protein was also detected in the case of posttreatment with ACHP in both HDFa and HLFa. This is of interest, as post-treatment with ACHP reduced aSMA in HDFa only, not in HLFa. Thus, ACHP is not only able to inhibit or reverse myofibroblast formation; it is also able to inhibit collagen synthesis, even in cells that still show myofibroblast properties. The HLFa data show that a downregulation in collagen type I synthesis is not automatically accompanied by a reduction in aSMA expression. These data are consistent with the opposite observation, namely that myofibroblasts which are negative for aSMA are still capable of synthesizing large quantities of collagen [31].
The collagen in fibrotic tissue has a higher level of pyridinoline cross-links compared to collagen in normal tissue [32][33][34]. This is because of the higher level of lysyl hydroxylation in the telopeptides of collagen. There are three LHs that convert lysine into hydroxylysine: LH1, LH2 (encompassing the splice variants LH2a and LH2b) and LH3. The telopeptides are hydroxylated by LH2b, whereas LH1 and LH3 hydroxylate certain Lys residues in the triple helical part of the collagen molecule [35]. The substrate specificity of LH2a is unknown, but most likely not the telopeptides. A higher expression of LH2 has been reported in fibrosis (specifically: LH2b) [32][33][34]36], explaining the higher level of pyridinoline cross-links. How the expression of PLOD1, PLOD2 and PLOD3 (encoding LH1, LH2 and LH3 respetively) is regulated is essentially unknown. We therefore investigated the effect of ACHP on expression levels of the three LHs. ACHP inhibited the expression of all three PLODs in HDFa, whereas no such inhibition was seen in HLFa. It thus seems the regulatory pathways that result in the expression of the three LHs differs, at least in part, between HDFa and HLFa.
In conclusion, the main message of this study is that ACHP exhibit strong anti-fibrotic properties. It is able to directly abolish the TGFb1induced myofibroblast formation and the synthesis of ECM components collagen type I and fibronectin. It also has the ability to partly reverse existing myofibroblasts into fibroblasts and the subsequent deposition of collagen type I. By inhibiting IKK, the NF-jB/IjB complex remains in the cytoplasm, preventing the translocation of NF-jB to the nucleus. The data obtained with ACHP shows that the NF-jB pathway play an important role in promoting fibrosis.