bone marrow mononuclear cell transplantation increases metalloproteinase-9 and 13 and decreases...

E-Mail [email protected]

Original Paper

Cells Tissues Organs DOI: 10.1159/000353215

Bone Marrow Mononuclear Cell Transplantation Increases Metalloproteinase-9 and 13 andDecreases Tissue Inhibitors of Metalloproteinase-1and 2 Expression in the Liver of Cholestatic Rats

Simone Nunes de Carvalho a Edward Helal-Neto b Daniela Caldas de Andrade a

Erika Afonso Costa Cortez a Alessandra Alves Thole a Christina Barja-Fidalgo b

Lais de Carvalho a a

Laboratório Cultura de Células, Departamento de Histologia e Embriologia, and b Laboratório de Farmacologia

Celular e Molecular, Departamento de Biologia Celular, Instituto de Biologia, Universidade do Estado doRio de Janeiro, UERJ, Rio de Janeiro , Brazil

assessed by Western blotting, along with α-SMA, CD68 and CD11b expression by confocal microscopy. Western blotting analysis showed that 14-day BDL animals had significantly reduced amounts of MMP-2 and MMP-13, but increased amounts of MMP-9 compared to normal rats. After 21 days of BDL, overall MMP amounts were decreased and TIMPs were increased. BMMNC transplantation significantly in-creased MMP-9 and MMP-13, and decreased TIMP expres-sion. Increased MMP activity was confirmed by zymography. MMP-9 and MMP-13 were expressed by macrophages near fibrotic septa, suggesting BMMNC may stimulate MMP pro-duction in fibrotic livers, contributing to ECM degradation and hepatic regeneration. © 2013 S. Karger AG, Basel

Introduction

Liver fibrosis is a pathological process caused by a chronic injury to the hepatic parenchyma, which can lead to cirrhosis and hepatic failure, a severe condition with high morbidity and mortality rates. There are several well-established factors that contribute to liver injury, among them hepatotoxic chemicals (CCl 4 , acetamino-

Key Words

Liver fibrosis · Bone marrow cells · Extracellular matrix · Metalloproteinases · Hepatic regeneration

Abstract

Liver fibrosis results from chronic injury followed by activa-tion of macrophages and fibrogenic cells like myofibroblasts and activated hepatic stellate cells. These fibrogenic cells ex-press α-smooth muscle actin (α-SMA) and produce excessive extracellular matrix (ECM), with disorganization and loss of function of hepatic parenchyma. It is known that increased levels of metalloproteinases (MMPs) in liver fibrosis are as-sociated with reduction of the pathologic ECM and fibrosis resolution. Recently, it has been shown that bone marrow mononuclear cells (BMMNCs) may reduce collagen and α-SMA expression, and ameliorate liver function in chole-static rats. Therefore, this study aimed to analyze MMP-2, MMP-9 and MMP-13, and tissue inhibitors of MMPs (TIMPs)-1 and TIMP-2 in the liver of cholestatic rats transplanted with BMMNC. Animals were divided into normal rats, cholestatic rats obtained after 14 and 21 days of bile duct ligation (BDL), and rats obtained after 14 days of BDL that received BMMNCs and were killed after 7 days. MMP and TIMP expression was

Accepted after revision: May 21, 2013 Published online: July 24, 2013

Dr. Lais de Carvalho Laboratório Cultura de Células, Departamento de Histologia e Embriologia Instituto de Biologia, Universidade do Estado do Rio de Janeiro, UERJ Av. Prof. Manoel de Abreu 444, 3° andar, Rio de Janeiro 20550-170 (Brazil) E-Mail ldc29 @ globo.com

© 2013 S. Karger AG, Basel 1422–6405/13/0000–0000$38.00/0

www.karger.com/cto

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM

Nunes de Carvalho et al.

Cells Tissues OrgansDOI: 10.1159/000353215

2

phen, ethanol and other drugs), congenital conditions (biliary atresia, inborn errors of metabolism), hepatitis viruses and hepatobiliary tumors [Bataller and Brenner, 2005; Hirschfield and Gershwin, 2011].

The result of a chronic, continued lesion to the liver is the activation of inflammatory cells that release cytokines and other factors that can, in turn, activate portal fibro-blasts and perisinusoidal hepatic stellate cells (HSCs) to differentiate into myofibroblasts and activated HSCs, re-spectively. The later cells have a fibrogenic phenotype, with synthesis and accumulation of large amounts of ex-tracellular matrix (ECM) proteins, and are characterized by α-smooth muscle actin (α-SMA) expression. The pathologic ECM accumulates in perisinusoidal and portal spaces, impairing hepatocyte nutrition and leading to dis-organization of the hepatic parenchyma that hallmarks liver fibrosis. The damaged liver progressively loses its metabolic capacity due to hepatocyte apoptosis and/or necrosis [Henderson and Iredale, 2007; Guicciardi and Gores, 2010; Forbes and Parola, 2011].

In the normal liver, ECM composition is closely con-trolled by groups of enzymes that can promote ECM deg-radation, such as matrix metalloproteinases (MMPs), or inhibit the action of MMPs and in turn promote ECM deposition, such as tissue inhibitors of MMPs (TIMPs). The physiological balance between MMPs and TIMPs is of most relevance to maintain the normal hepatic cytoar-chitecture. However, in the injured liver, the inflamma-tory process implies disturbances in this balance, with predominance of ECM deposition and fibrosis develop-ment [Consolo et al., 2009; Forbes and Parola, 2011; Ra-machandran and Iredale, 2012].

In both normal and pathological conditions, inflam-matory cytokines such as tumor necrosis factor-α (TNF-α) and interleukin-1 (IL-1) are the main orchestrators of

MMP expression in the liver. However, a single cytokine may stimulate a certain MMP in one cell type but inhibit the same MMP in another, and thus analysis of MMP modulation is complex. Besides cytokines, defined ECM components as well as mechanical stress also determine the expression of MMP [Han, 2006].

Different cell populations within the liver contribute to MMP synthesis. Inflammatory cells such as Kupffer cells and neutrophils produce large amounts of different MMPs in pathological conditions. Interestingly, hepato-cytes were found to produce some matrix MMPs (MMP-2, MMP-3, MMP-10 and MMP-13) and two TIMPs (TIMP-1 and TIMP-2) [Garciade Leon Mdel et al., 2006].

After a liver injury, Kupffer cells are stimulated to release high amounts of MMP-9, which is specific of hepatic in-flammatory response, causing localized ECM degradation. Other sources of MMPs in the injured liver are fibrogenic α-SMA + cells, which produce MMP-2, MMP-9 and MMP-13 in varying amounts. However, in the later phases of fi-brosis progression, MMPs become inactivated by the in-creasing expression of TIMPs, leading to an excessive ECM accumulation. The disruption of ECM balance towards a fibrogenic scenario is the hallmark of liver fibrosis, and once initiated, this process can only be reverted by the inter-ruption of liver injury [Arthur, 2000; Iredale, 2007; Forbes and Parola, 2011]. The major producers of TIMPs in he-patic fibrosis are fibrogenic cells within the liver, and these cells are maintained by its own secreted products like col-lagen, fibronectin and other ECM molecules, along with growth factors produced by inflammatory cells such as transforming growth factor-β (TGF-β) and platelet-derived growth factor (PDGF) [Arthur, 2000; Clark et al., 2008; Consolo et al., 2009]. It is known that liver regeneration is achieved once the inflammatory response is ceased or re-duced, resulting in cell cycle arrest of fibrogenic cells, which can either enter an apoptotic process or become quiescent cells, reducing ECM accumulation [Guyot et al., 2006; Forbes and Parola, 2011; Ramachandran and Iredale, 2012].

Currently, there are few available therapies to ad-vanced liver fibrosis, known as cirrhosis, and liver trans-plantation is the only effective treatment in the most se-vere cases. However, many patients suffer from organ shortage, which has prompted the search for alternative methods to delay or stop fibrosis progression. In this re-spect, bone marrow mononuclear cell (BMMNC) trans-plantation has been used in experimental and clinical protocols to stimulate liver regeneration and fibrosis re-gression [Higashiyama et al., 2007; Gilchrist and Plevris, 2010; Kisseleva et al., 2010; Inagaki et al., 2012]. Recent reports show BMMNC therapy is associated with relevant

Abbreviations used in this paper

α-SMA α-smooth muscle actin AU BDL BMMNCs ECM HSCs IL-1 MMPs PDGF TBS TIMPs TNF-α

arbitrary units bile duct ligation bone marrow mononuclear cells extracellular matrix hepatic stellate cells interleukin-1 metalloproteinases platelet-derived growth factor Tris-buffered saline tissue inhibitors of metalloproteinases tumor necrosis factor-α

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM

Liver Fibrosis and Cell Therapy Cells Tissues OrgansDOI: 10.1159/000353215

3

benefits to hepatic function and a significant decrease in collagen and α-SMA expression [Carvalho et al., 2010].

The aim of this work was to evaluate MMP expression and activity and its balance with TIMP expression within the liver of fibrotic rats transplanted with BMMNC, and to analyze MMP expression by the two major cell types responsible for its regulation – macrophages and fibro-genic α-SMA + cells – in response to cell therapy. Chole-static rats were obtained after bile duct ligation (BDL), and transplanted with BMMNC on the 14th day after BDL. After 7 days, livers were analyzed for MMP-2, MMP-9 and MMP-13, and TIMP-1 and TIMP-2 content by Western blotting, MMP activity was accessed by zymography blots, and MMP-producing cells directly involved in fibrosis progression were identified by confocal microscopy.

Materials and Methods

Hepatic Fibrosis Induction and Experimental Groups All animals used in this study were submitted to protocols ap-

proved by the Animal Experiments Committee in accordance with the standard guidelines for animal experimentation, and received water and standard food ad libitum. To induce hepatic fibrosis, 3-month-old male Wistar rats (250–270 g) were put under halo-thane inhalation anesthesia and BDL surgery followed by resection was carried out. BMMNCs were isolated from the tibias and fe-murs of 2-month-old healthy male Wistar rats, sacrificed in a CO 2 chamber. The medullary cavities of the bones were exposed and flushed with cold Dulbecco’s Modified Eagle’s Medium (Sigma-Aldrich), pH 7.2. The bone marrow cell were submitted to a Ficoll-Hypaque (Sigma-Aldrich) density gradient and, after centrifuga-tion at 2,000 rpm, the mononuclear cell interface was collected and transplanted via jugular veins in fibrotic rats obtained after 14 days of BDL surgery.

The animals were divided into 4 groups (n = 6 per group): normal animals, animals with hepatic fibrosis after 14 and 21 days of BDL, and animals with hepatic fibrosis after 14 days of BDL that received 1 × 10 7 BMMNC cells via jugular vein and were sacrificed after 7 days of transplantation. All animals were killed in a CO 2 chamber.

Western Blotting Analysis of MMPs and TIMPs Liver fragments (100 mg) from experimental groups were in-

cubated in lysis buffer (1% Triton X-100, 100 m M Tris, 100 m M sodium pyrophosphate, 10 m M EDTA, 10 m M sodium orthovana-date and the protease inhibitors PMSF and aprotinin) and me-chanically dissociated before sonication. After centrifugation (10,000 rpm, 10 min), supernatant was collected and proteins de-termined by the BCA method (Thermo Scientific kit). After the addition of sample buffer, samples were maintained at –20 ° C un-til analysis. Protein electrophoresis was performed in 12% poly-acrylamide gel (1 h, 150 V) and proteins were transferred to PVDF membranes (1 h, 15 V). The membranes were blocked overnight in phosphate-buffered saline containing 0.1% Tween 20 (Tris-buffered saline) and 5% albumin from bovine serum. Then, mem-branes were incubated with anti-rat MMP-2 (sc-10736), MMP-9 (sc-10737), MMP-13 (sc-30073), TIMP-1 (sc-5538; all Santa Cruz

Biotechnology), TIMP-2 (F27P3A4; BioLegend) or β-actin (A2066; Sigma-Aldrich) polyclonal primary antibodies 1: 500 in TBS, fol-lowed by washing and incubation with biotinylated anti-goat sec-ondary antibody (Life Technologies) and finally streptavidin-per-oxidase (Life Technologies), both 1: 5,000 in TBS. Membranes were then incubated with substrate for enhanced chemiluminescence containing hydrogen peroxide (Thermo Scientific) for 5 min and revealed in photographic films (Kodak) in a dark chamber. The bands were then scanned and densitometry of digitalized images was performed using Adobe Photoshop CS6 software. Values are expressed as arbitrary units (AU). All proteins were normalized to the internal control, β-actin, during analysis.

Zymography Liver fragments (100 mg) from experimental groups were in-

cubated in zymography lysis buffer (1% Triton X-100, 100 m M Tris and 100 m M protease inhibitor PMSF) and mechanically dissoci-ated. After centrifugation (10,000 rpm, 10 min), supernatant was collected and proteins were determined by the BCA method (Thermo Scientific kit). Sample buffer without β-mercaptoethanol was added and samples were maintained at –20 ° C until protein electrophoresis in 8% polyacrylamide gel (1 h, 150 V) containing 1 mg/ml porcine gelatin (Sigma-Aldrich). The gels were exten-sively washed in 2.5% Triton X-100 (Merck) and incubated in de-veloping buffer (50 m M Tris, 0.2 M NaCl, 5 m M CaCl 2 and 2 μ M ZnCl 2 ) at 37 ° C for 48 h. After staining with 0.25% Coomassie blue (Sigma-Aldrich) for 30 h, gel images were obtained with Chemidoc (BioRad) equipment and analyzed using Adobe Photoshop CS6 software. Values are expressed as AU.

Double Immunofluorescence The tissue was embedded in Tissue Tek and maintained at

–70 ° C until sectioning at 5–10 μm. Liver sections were then fixed in cold acetone and double immunofluorescence was proceeded by incubation with either monoclonal mouse anti-rat α-SMA (sc-53015; Santa Cruz Biotechnology), anti-CD68 (MCA341R; Sero-tec) or anti-CD11b (PE-conjugated; Caltag) primary antibodies along with polyclonal anti-rat MMP-2, MMP-9 or MMP-13 (San-ta Cruz Biotechnology) primary antibodies. Sections were thenincubated with anti-mouse Alexa-488-conjugated and Alexa-555-conjugated secondary antibodies (Life Technologies), fol-lowed by addition of the nuclear dye DAPI (Sigma-Aldrich). Slides were mounted and observed with a Zeiss LSM 510 META confocal laser scanning microscope.

Statistics Data were analyzed using one-way analysis of variance with

Tukey’s post hoc test for multiple comparisons, with p ≤ 0.05 being considered statistically significant. Data are presented as mean ± SE.

Results

Western Blotting Analysis of MMPs and TIMPs Results of Western blotting ( fig. 1 ) showed that animals

with 14-day BDL had significantly reduced amounts of MMP-2 (1,706.15 ± 488.44) and MMP-13 (34.76 ± 13.21) in comparison to normal rats (3,638.42 ± 1,069.12 and

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM



4

TIMP-1MMP-2

MMP-9 TIMP-2

-Actin

0

1,000

2,000

3,000

4,000

5,000AU

Normal

F14d

+ BM

MNC 7dF21d

F14d

*****

0

1

2

3

AU

Normal

F14d

+ BM

MNC 7dF21d

F14d

***

0

1,000

2,000

3,000

AU

Normal

F14d

+ BM

MNC 7dF21d

F14d

***** *

0

5

10

15

AU

Normal

F14d

+ BM

MNC 7dF21d

F14d

**

0

50

100

150

Normal

F14d

+ BM

MNC 7dF21d

F14d

MMP-13

AU

****

Fig. 1. Results of Western blotting analysis of MMP-2, MMP-9 and MMP-13, and TIMP-1 and TIMP-2 content, showing overall in-crease in MMP expression and decrease in TIMP-1 and TIMP-2

expression after BMMNC transplantation. The 4 respective groups are listed on the x-axes: F = fibrosis; d = days. * p ≤ 0.05; ** p ≤ 0.005; *** p ≤ 0.001.

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM


5

108.68 ± 18.62, respectively). We also observed that after 14 days of BDL, there were increased amounts of MMP-9 (270.28 ± 38.86) compared to normal rats (130.18 ± 28.92), probably due to the inflammatory activity of Kupffer cells. However, the BMMNC group expressed significantly higher amounts of MMP-9 (166.76 ± 39.48) and MMP-13

(42.10 ± 12.52) compared to the 21-day BDL group (70.73 ± 12.26 and 15.32 ± 6.22, respectively), but not of MMP-2.

TIMP-1 expression was significantly increased in the 14-day BDL (2.47 ± 0.26) and in the 21-day BDL groups (2.20 ± 0.3) when compared to the normal group (0.57 ± 0.2), but not in the BMMNC group (1.28 ± 0.36). TIMP-2

82 kDaActiveMMP-9

0

1

2

3

4

AU

Normal

F14d

+ BM

MNC 7dF21d

F14d

***

a b20 μm 20 μm

Fig. 3. a , b Confocal microscopy images of immunofluorescence results using anti-α-SMA (green) and anti-MMP-2 (red) primary antibodies, performed in liver sections of the BMMNC-transplanted group. Nuclei were stained with DAPI (blue). MMP-2 expression was observed in areas of hepatic disorganization, near α-SMA-expressing areas (arrows).

Fig. 2. Zymography results. Bands corre-spond to active MMP-9 (82 kDa). Fibrotic groups of 14-day and 21-day BDL showed decreased MMP activity when compared to normal and BMMNC-transplanted groups. * p ≤ 0.05; ** p ≤ 0.005.

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM



6

expression was significantly increased in the 21-day BDL group (10.81 ± 4.70) compared to the normal (4.19 ± 2.98) and 14-day BDL (5.29 ± 1.91) groups. BMMNC transplantation significantly decreased TIMP-2 amounts (4.09 ± 1.58), which were similar to normal livers.

Zymography Liver zymography of gelatin-containing blots showed

MMP activity in different molecular weights, as expected for this tissue, particularly in the normal group ( fig. 2 ).

The active fragment of MMP-9 (82 kDa) was considered for analysis based on Western blotting results and feasi-bility. Results showed that overall MMP activity was higher in the normal group, particularly MMP-9 (2.92 ± 0.29), and was significantly decreased in fibrotic groups of 14-day BDL (0.78 ± 0.13) and 21-day BDL (0.61 ± 0.13). BMMNC transplantation significantly augmented MMP-9 activity (1.68 ± 0.05) compared to fibrotic groups.

a b

c d

20 μm 20 μm

20 μm 20 μm

Fig. 4. Confocal microscopy images of immunofluorescence results using anti-CD68 (green) and anti-MMP-9 (red) primary antibodies. Nuclei were stained with DAPI (blue). a A liver section of the 21-day BDL group, pre-senting little MMP-9 expression. b–d Liver sections of the BMMNC-transplanted group. MMP-9 expression is observed near CD68-expressing areas.

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM


7

a b20 μm 20 μm

a b20 μm 20 μm

Fig. 5. Confocal microscopy images of immunofluorescence results using anti-CD68 (green) and anti-MMP-13 (red) primary antibodies. Nuclei were stained with DAPI (blue). a A liver section of the 21-day BDL group, pre-senting little MMP-13 expression. b A liver section of BMMNC-transplanted group, with increased MMP-13 expression (arrows). Colocalization with CD68+ cells was absent.

Fig. 6. Confocal microscopy images of immunofluorescence results using anti-CD11b (red) and anti-MMP-13 (green) primary antibodies. Nuclei were stained with DAPI (blue). a A liver section of the 21-day BDL group, presenting little MMP-13 expression, colocalized with numerous cells expressing the granulocyte/macrophage marker CD11b in fibrous septa (arrows). Cells display macrophage morphology. b A liver section of the BMMNC-transplanted group, with high MMP-13 expression in CD11b+ cells in a fibrous septum.

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM



8

Double Immunofluorescence Confocal microcopy results showed higher labeling

to MMP-2, MMP-9 and MMP-13 in the BMMNC-transplanted group when compared to the 21-day BDL group ( fig. 3–6 ), confirming Western blotting data. We also observed MMP-2 expression mainly near α-SMA + areas in fibrotic septa in the BMMNC-transplanted group ( fig. 3 ), suggesting fibrotic cells may be respon-sible for most of the MMP-2 production. Additionally, MMP-9 expression was observed mainly by Kupffer cells also in fibrotic areas ( fig. 4 ) of both the 21-day BDL and BMMNC-transplanted groups. Although MMP-13 expression was again observed in fibrotic regions, CD68+ macrophages did not appear to be the main cell source of this protein ( fig. 5 ). Instead, CD11b+ cells ap-peared to be the main MMP-13-producing cell typein our model ( fig. 6 ), in both experimental groups ob-served. Furthermore, livers from 21-day BDL animals displayed more CD11b+ cells than BMMNC-trans-planted livers. Taken together, these results show that BMMNC may stimulate MMP production by different cell types in fibrotic livers, contributing to ECM degra-dation and hepatic regeneration.

Discussion

MMP expression and activity in cholestatic rats trans-planted with BMMNCs was analyzed along with the ma-jor cell types involved in MMP synthesis within the liver for the first time. Our group has previously shown that BMMNCs can reduce fibrosis and ameliorate hepatic function by significantly decreasing collagen and α-SMA expression [Carvalho et al., 2010]. However, there are many questions regarding the mechanisms responsible for the hepatic regeneration induced by BMMNCs. It is known that fibrosis resolution is associated with an in-crease in MMP expression, which can in turn stimulate ECM remodeling, with pathologic ECM degradation and reestablishment of the normal hepatic parenchyma. Therefore, MMP expression and ECM balance may be directly or indirectly affected by BMMNC transplanta-tion in rats with liver fibrosis, as observed in other studies [Higashiyama et al., 2007]. Indeed, we observed that BMMNCs can significantly increase MMP-9 and MMP-13 expression in fibrotic livers. This leads to collagen deg-radation and loss of focal contacts by fibrogenic cells, which can enter apoptosis or become quiescent. It is known that the MMP-mediated resolution of tissue fibro-sis may occur through ECM degradation as well as by in-

duction of fibrogenic cell apoptosis. The consequence is a significant reduction in ECM synthesis by these cells, favoring hepatic regeneration [Han, 2006].

MMP sources within the liver may include quiescent or activated HSCs, fibroblasts or myofibroblasts, quies-cent and activated HSCs and Kupffer cells. The pattern of MMP expression varies according to the respective cell type. Kupffer cells produce mainly MMP-9, and some MMP-2 and MMP-13 in minor amounts, and MMP-9 expression by these cells increases after liverinjury and inflammatory stimuli [Arthur, 2000]. In the injured hepatic microenvironment, MMP-9 can cleave and activate growth factors like TGF-β and PDGF, cru-cial to the activation of fibrogenic cells [Consolo et al., 2009]. This may explain why soon after MMP-9 increase caused by liver damage, an apparently paradoxical reac-tion occurs with increase in ECM deposition, although many other matrix molecules, growth factors and cyto-kines contribute to TGF-β and PDGF expression and activation.

In the later phases of fibrosis, MMP-9 expression and activity decreases dramatically, with the consolidation of the fibrotic process, as we observed in the group with 21 days of BDL. This is in accordance with previous reports showing that established fibrosis courses with a wide-spread decrease in MMP expression in the liver [Knittel et al., 2000; Glaser et al., 2009]. However, BMMNC trans-plantation significantly increased MMP-9 expression andactivity compared to the 21-day BDL group. MMP-9 is known to play a role in hematopoietic stem cell detach-ment from the bone marrow niche and migration to the blood stream. These cells can then reach the injured liver, attracted by chemotactic factors released by the organ such as SDF-1 [Higashiyama et al., 2007]. Therefore, the increased amounts of MMP-9 produced in the liver after a hepatic lesion can reach the bone marrow through blood circulation, and promote the migration of endogen hematopoietic stem cells that may be beneficial to liver disease.

Confocal microscopy analysis showed that MMP-9 ex-pression is observed mainly in CD68+ Kupffer cells in the BMMNC group, showing that Kupffer cells may play a role in this model of hepatic regeneration. A recent study characterizing macrophage populations in hepatitis C subjects has shown that the MMP-9-producing cells were observed mainly near portal areas rather than hepatic si-nusoids [Gadd et al., 2013]. In the present model of BDL, CD68+ cells expressing MMP-9 were also seen preferen-tially near portal spaces and fibrous septa in both fibro-genic and regenerating processes. These findings support

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM


9

the recent hypothesis that liver macrophages consist ofa heterogeneous population of cells with different rolesin pathophysiological conditions of the organ. In fact, Kupffer cells have recently been shown to be key cells in liver health and disease, participating in either the initia-tion of the fibrotic process, by expression of an inflamma-tory/fibrogenic pattern of cytokines and growth factors such as IL-6, TNF-α, TGF-β and PGDF, or in fibrosis re-gression and hepatic regeneration, by release of anti-in-flammatory cytokines such as IL-10 and antifibrogenic factors such as MMPs [Heymann et al., 2009; Wynn and Barron, 2010]. In this regard, the bone marrow mono-nuclear fraction contains hematopoietic stem cells and monocytes in the differentiation process, and we hypoth-esize that these cells can be precursors of Kupffer cells within the damaged liver. Although this hypothesis was not directly addressed in this study, it is well known that bone marrow-derived monocytes are the precursors of Kupffer cells in many pathophysiological conditions in the liver [Karlmark et al., 2009]. Additionally, other cells of the mononuclear fraction, such as lymphocytes, can in turn stimulate an anti-inflammatory response by Kupffer cells once they encounter an altered hepatic microenvi-ronment.

Studies using a model of bile duct recanalization after BDL showed that the regenerating liver presents massive apoptosis of cytokeratin 19 + cholangiocytes, which are colocalized with CD68+ macrophages, accompanied by a significant increase in MMP activity [Popov et al., 2010]. Our previous work has also shown that BMMNC-treated animals present fewer cytokeratin 19 + cholangiocytes than the 21-day BDL group [Carvalho et al., 2010], thus indicating that macrophages may participate in both phagocytosis of cholangiocyte apoptotic bodies as well as in tissue remodeling through MMP production during liver regeneration in cholestatic diseases.

In the normal liver, perisinusoidal HSCs and portal fibroblasts are the main producers of MMP-2, along with Kupffer cells. This expression is also seen in fibrogenic α-SMA + cells (activated HSCs and myofibroblasts), al-though TIMP-1 and TIMP-2, also produced in large amounts by these cells, inhibit the overall MMP activity, leading to ECM deposition [Arthur, 2000; Consolo et al., 2009]. Western blotting analysis showed a significant re-duction in MMP-2 expression in the fibrotic liver of 14- and 21-day BDL rats, along with increased TIMP-1 and TIMP-2 expression when compared to normal livers, which is expected once fibrogenic cells proliferate after BDL. BMMNC transplantation did not significantly alter MMP-2 expression in comparison with rats with 21 days

of BDL. We have reported that BMMNCs reduce α-SMA expression in fibrotic livers [Carvalho et al., 2010] leading to fibrogenic cell apoptosis or quiescence [Nunes de Car-valho et al., 2013]. Therefore, we suggest that since the major producers of MMP-2 are fibrogenic cells, their de-crease may explain the unaltered level of MMP-2 expres-sion in the liver of BMMNC-transplanted rats. Confocal microscopy results confirmed MMP-2 expression by fi-brogenic and Kupffer cells within the liver of both 21-day BDL and BMMNC-treated rats.

Recent reports have shown that MMP-13 may be im-portant to ECM remodeling and fibrosis regression, after its expression was often increased in a number of rodent models of liver fibrosis where hepatic regeneration was stimulated [Consolo et al., 2009; Endo et al., 2011; Forbes and Parola, 2011]. MMP-1 is the human analogue of MMP-13, and is also correlated with liver regeneration in clinical protocols [Arthur, 2000]. However, besides the well-established role of MMP-13 in matrix remodeling in fibrotic livers, it is also thought to participate in initial liver fibrogenesis, as shown in MMP-13-deficient mice [Uchinami et al., 2006]. Regarding the scarce available data on the cell type responsible for MMP-13 production in the liver, some authors have shown that macrophages and HSCs [Altadill et al., 2009; Ohyama et al., 2011] may produce MMP-13 in the liver, but other cell types could be involved. In the present study, we observed for the first time a significant increase in MMP-13 expression, shown to colocalize with CD11b+ cells spread through fibrous septa in the liver of cholestatic rats transplanted with BMMNC. The CD11b marker is expressed by mono-cytes/macrophages as well as granulocytes, although there is little evidence that neutrophils express MMP-13. Recent work demonstrates that CD68+ cells may be liver-resident macrophages (Kupffer cells), and CD11b+ cells may represent resident as well as transient cells in injured livers [Nakashima et al., 2013]. Therefore, we suggest that these particular MMP-13-expressing cells may corre-spond to a specific subset of liver macrophages, different from the MMP-9-expressing CD68+ population, which did not express MMP-13 in our model. These data con-firm not only the findings of the role of MMP-13 in ECM remodeling during hepatic fibrosis regression, but also the contribution of diverse macrophage populations to this process.

In conclusion, BMMNC transplantation is associated with a significant increase in MMP-9 and MMP-13 ex-pression, accompanied by a decrease in TIMP-1 and TIMP-2 expression in the liver of cholestatic rats, which represents a highly antifibrogenic scenario that favors

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM



10

ECM remodeling in liver fibrosis. Additionally, MMP-9 and MMP-13 were expressed mainly by macrophages in fibrotic septa of the regenerating liver, and we suggest that BMMNCs may contribute to this cell population, which may play an important role in the present model of liver fibrosis and regeneration after cell therapy.

Acknowledgements

This study was supported by Conselho Nacional de Desen-volvimento Científico e Tecnológico (CNPq), Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ), and Coordenação de Aperfeiçoamento de Pessoal de nível Superior (Capes).

References

Altadill, A., M. Rodriguez, L.O. Gonzalez, S. Jun-quera, M.D. Corte, M.L. Gonzalez-Dieguez, A. Linares, E. Barbon, M. Fresno-Forcelledo, L. Rodrigo, F.J. Vizoso (2009) Liver expres-sion of matrix metalloproteases and their in-hibitors in hepatocellular carcinoma. Dig Liv-er Dis 41: 740–748.

Arthur, M.J. (2000) Fibrogenesis II. Metallopro-teinases and their inhibitors in liver fibrosis. Am J Physiol Gastrointest Liver Physiol 279: G245–G249.

Bataller, R., D.A. Brenner (2005) Liver fibrosis. J Clin Invest 115: 209–218.

Carvalho, S.N., D.C. Lira, G.P. Oliveira, A.A. Thole, A.C. Stumbo, C.E. Caetano, R.G. Marques, L. Carvalho (2010) Decreased col-lagen types I and IV, laminin, CK-19 and al-pha-SMA expression after bone marrow cell transplantation in rats with liver fibrosis. His-tochem Cell Biol 134: 493–502.

Clark, I.M., T.E. Swingler, C.L. Sampieri, D.R. Ed-wards (2008) The regulation of matrix metal-loproteinases and their inhibitors. Int J Bio-chem Cell Biol 40: 1362–1378.

Consolo, M., A. Amoroso, D.A. Spandidos, M.C. Mazzarino (2009) Matrix metalloproteinases and their inhibitors as markers of inflamma-tion and fibrosis in chronic liver disease (re-view). Int J Mol Med 24: 143–152.

Endo, H., M. Niioka, Y. Sugioka, J. Itoh, K. Ka-meyama, I. Okazaki, R. Ala-Aho, V.M. Kaha-ri, T. Watanabe (2011) Matrix metallopro-teinase-13 promotes recovery from experi-mental liver cirrhosis in rats. Pathobiology 78: 239–252.

Forbes, S.J., M. Parola (2011) Liver fibrogenic cells. Best Pract Res Clin Gastroenterol 25: 207–217.

Garciade Leon Mdel, C., I. Montfort, E. Tello Montes, R. Lopez Vancell, A. Olivos Garcia, A. Gonzalez Canto, M. Nequiz-Avendano, R. Perez-Tamayo (2006) Hepatocyte production of modulators of extracellular liver matrix in normal and cirrhotic rat liver. Exp Mol Pathol 80: 97–108.

Gadd, V.L., M. Melino, S. Roy, L. Horsfall, P. O’Rourke, M.R. Williams, K.M. Irvine, M.J. Sweet, J.R. Jonsson, A.D. Clouston, E.E. Pow-ell (2013) Portal, but not lobular, macro-phages express matrix metalloproteinase-9: association with the ductular reaction and fi-brosis in chronic hepatitis C. Liver Int 33: 569–579.

Gilchrist, E.S., J.N. Plevris (2010) Bone marrow-derived stem cells in liver repair: 10 years down the line. Liver Transpl 16: 118–129.

Glaser, S.S., E. Gaudio, T. Miller, D. Alvaro, G. Alpini (2009) Cholangiocyte proliferation and liver fibrosis. Expert Rev Mol Med 11: e7.

Guicciardi, M.E., G.J. Gores (2010) Apoptosis as a mechanism for liver disease progression.Semin Liver Dis 30: 402–410.

Guyot, C., S. Lepreux, C. Combe, E. Doudnikoff, P. Bioulac-Sage, C. Balabaud, A. Desmouliere (2006) Hepatic fibrosis and cirrhosis: the (myo)fibroblastic cell subpopulations in-volved. Int J Biochem Cell Biol 38: 135–151.

Han, Y.P. (2006) Matrix metalloproteinases, the pros and cons, in liver fibrosis. J Gastroen-terol Hepatol 21(suppl 3): S88–S91.

Henderson, N.C., J.P. Iredale (2007) Liver fibro-sis: cellular mechanisms of progression and resolution. Clin Sci (Lond) 112: 265–280.

Heymann, F., C. Trautwein, F. Tacke (2009) Monocytes and macrophages as cellular tar-gets in liver fibrosis. Inflamm Allergy Drug Targets 8: 307–318.

Higashiyama, R., Y. Inagaki, Y.Y. Hong, M. Ku-shida, S. Nakao, M. Niioka, T. Watanabe, H. Okano, Y. Matsuzaki, G. Shiota, I. Okazaki (2007) Bone marrow-derived cells express matrix metalloproteinases and contribute to regression of liver fibrosis in mice. Hepatol-ogy 45: 213–222.

Hirschfield, G.M., M.E. Gershwin (2011) Primary biliary cirrhosis: one disease with many faces. Isr Med Assoc J 13: 55–59.

Inagaki, Y., R. Higashiyama, K. Higashi (2012) Novel anti-fibrotic modalities for liver fibro-sis: molecular targeting and regenerative medicine in fibrosis therapy. J Gastroenterol Hepatol 27(suppl 2): 85–88.

Iredale, J.P. (2007) Models of liver fibrosis: ex-ploring the dynamic nature of inflammation and repair in a solid organ. J Clin Invest 117: 539–548.

Karlmark, K.R., R. Weiskirchen, H.W. Zimmer-mann, N. Gassler, F. Ginhoux, C. Weber, M. Merad, T. Luedde, C. Trautwein, F. Tacke (2009) Hepatic recruitment of the inflamma-tory Gr1+ monocyte subset upon liver injury promotes hepatic fibrosis. Hepatology 50: 261–274.

Kisseleva, T., E. Gigante, D.A. Brenner (2010) Re-cent advances in liver stem cell therapy. Curr Opin Gastroenterol 26: 395–402.

Knittel, T., M. Mehde, A. Grundmann, B. Saile, J.G. Scharf, G. Ramadori (2000) Expression of matrix metalloproteinases and their inhibi-tors during hepatic tissue repair in the rat. Histochem Cell Biol 113: 443–453.

Nakashima, H., Y. Ogawa, S. Shono, M. Kinoshi-ta, M. Nakashima, A. Sato, M. Ikarashi, S. Seki (2013) Activation of CD11b + Kupffer cells/macrophages as a common cause for exacer-bation of TNF/Fas-ligand-dependent hepati-tis in hypercholesterolemic mice. PLoS One 8: e49339.

Nunes de Carvalho, S., D. da Cunha Lira, E.A. Costa Cortez, D.C. de Andrade, A.A. Thole, A.C. Stumbo, L. de Carvalho (2013) Bone marrow cell transplantation is associated with fibrogenic cells apoptosis during hepatic re-generation in cholestatic rats. Biochem Cell Biol 91: 88–94.

Ohyama, T., Y. Yamazaki, K. Sato, N. Horiguchi, T. Ichikawa, S. Kakizaki, H. Takagi, M. Mori (2011) Transforming growth factor-α attenu-ates hepatic fibrosis: possible involvement of matrix metalloproteinase-1. Liver Int 31: 572–584.

Popov, Y., D.Y. Sverdlov, K.R. Bhaskar, A.K. Sharma, G. Millonig, E. Patsenker, S. Krahen-buhl, L. Krahenbuhl, D. Schuppan (2010) Macrophage-mediated phagocytosis of apop-totic cholangiocytes contributes to reversal of experimental biliary fibrosis. Am J Physiol Gastrointest Liver Physiol 298: G323–G334.

Ramachandran, P., J.P. Iredale (2012) Liver fibro-sis: a bidirectional model of fibrogenesis and resolution. QJM 105: 813–817.

Uchinami, H., E. Seki, D.A. Brenner, J. D’Armiento (2006) Loss of MMP 13 attenuates murine he-patic injury and fibrosis during cholestasis. Hepatology 44: 420–429.

Wynn, T.A., L. Barron (2010) Macrophages: mas-ter regulators of inflammation and fibrosis. Semin Liver Dis 30: 245–257.

Dow

nloa

ded

by:

Lule

a T

ekni

ska

Uni

vers

itet

13

0.24

0.43

.43

- 8/

19/2

013

5:28

:32

PM

bone marrow mononuclear cell transplantation increases metalloproteinase-9 and 13 and decreases...

Documents