Bone Marrow Mononuclear Cells (BMMC) have emerged as potential candidates for cell therapy due to their ease of use. In models of acute liver failure, transplantation of these cells increase the survival rate [1,2]. Furthermore, recent clinical trials have demonstrated that transplantation of these cells or their fractions improves the condition of patients with cirrhosis [3-5], as well as improve liver function in animal models of cirrhosis [6]. Moreover, BMMC are characterized by their ability to differentiate into several functional mature cell types both in vivo and in vitro, including cardiomyocytes [7,8], endothelial cells [9], neurons [10,11] and hepatocytes [12-18]. In addition, many groups have developed protocols for the in vitro differentiation of bone marrow-derived cells into hepatocyte-like cells [19-24]. Although bone marrow mononuclear fraction is used for in vivo transplantation, in vitro protocols usually work with Mesenchymal Stem Cells (MSC) [19,23-25].

Microencapsulation provides a vehicle for the discrete control of key parameters such as the diffusion of growth factors, metabolites, and wastes. It has been demonstrated that the alginate microenvironment maintains cell viability, is conducive to embryonic stem cell differentiation into hepatocytes, and maintains differentiated cellular function [26]. In addition, encapsulated bone marrow cells showed evidence of glycogen synthesis and expression of typical markers of hepatocytes, after transplantation in the 90% liver failure model [15,27].

In the present work we used semi-permeable alginate microcapsules to isolate BMMC in a model of Carbon Tetrachloride (CCl₄)-induced acute liver injury. Alginate microcapsules isolate BMMC and allow the study of in vivo and in vitro differentiation through paracrine factors. We also focused on investigating early reprogramming events that might take place during exposure to injury microenvironment.

Experimental design

Both in vitro and in vivo studies were conducted in Wistar rats submitted to CCl₄-induced acute liver injury. For in vivo studies, encapsulated BMMC were transplanted into animals with or without liver injury by CCl₄ and kept for 48 hrs. For in vitro studies the encapsulated BMMC were co-cultured with hepatocytes isolated from rats with CCl₄-induced acute liver injury or without liver damage for 6 hrs (Figure 1).

Figure 1:Experimental design of in vivo and in vitro studies. † Sacrifice of animals and retrieve of capsules; Interruption of coculture and recovery of encapsulated BMMC

Animals

Adult male Wistar rats weighing 200 to 250 gr were kept under 24 hour light/dark cycles and fed standard chow and water ad libitum. Animals were sacrificed in the CO₂ chamber, either 6, 24 or 48 hrs after CCl₄ administration for in vivo studies (n=3/time point). For in vitro studies, animals (n=3) were submitted to hepatocyte isolation 24 hrs after CCl₄ administration. Control animals (n=12, 9 and 3 for in vivo and in vitro studies, respectively) were submitted to the same procedures but receiving vehicle instead of CCl₄. Bone marrow donors (n=9, 6 and 3 for in vivo and in vitro studies, respectively) were rats not submitted to any other procedure. This study was approved by the ethics research committee and national guidelines for animal care were followed.

Acute liver injury model

Acute liver injury was induced by a single CCl₄ (VETEC, Brazil) dose of 1.25 ml/kg diluted in olive oil [2]. A final volume of 1 ml was administered by gavage. Control animals received only olive oil by gavage.

Isolation of bone marrow mononuclear cells (BMMC)

BMMC were isolated from the femur and tibia of Wistar rats, as previously reported [1]. Briefly, bone marrow was flushed with Dulbecco's Modified Eagle Medium (DMEM-LGC, Brazil) supplemented with 10% Fetal Bovine Serum (FBS-Gibco, USA) and 1% Penicilin/Streptomicin (P/S-Gibco, USA) and BMMC were separated onto a Ficoll Histopaque (GE-Healthcare, USA) layer.

BMMC encapsulation

BMMC were encapsulated in sodium alginate microcapsules under sterile conditions, using the protocol described by our group [28,29]. BMMC were mixed with 1.5% sodium alginate (Sigma-Aldrich, USA) in DMEM and extruded through an encapsulation unit, type J1 (Nisco, Zurich, Switzerland), attached to a syringe pump (JMS, Singapore). Droplets were sheared off with an air flow of 5 L/min delivered to the tip of a 27-G needle and the rate of infusion was 40 mL/h. The droplets fell into a bath of 125 mM CaCl₂ and ionically cross-linked with Ca2⁺ to form solid spherical hydrogel beads containing embedded cells. For in vitro experiment, in each well, beads were produced from a volume of 100 µL of alginate suspension, containing 1x10⁵ of BMMC. The resulting beads were maintained under normal tissue culture conditions: DMEM supplemented with 10% FBS and 1% P/S at 37°C and 5% CO₂ for 24 hrs prior to administration. For in vivo experiment, in each well, beads were produced from a volume of 2 mL of alginate suspension, containing 1x10⁶ BMMC/animal.

Capsules transplantation

For in vivo experiments, 24 hrs after CCl₄ administration animals were anesthetized with inhaled isoflurane and a small incision was made in the abdomen. A total of 1x10⁶ encapsulated BMMC suspended in 2 mL saline buffer was placed in the peritoneal cavity (tCCl₄ group, n=3/time). The same procedure was performed in animals without liver injury (tControl group, n=3/time). Sacrifice was performed in CO₂ chambers 6, 24 or 48 hrs after capsules transplantation. Capsules were retrieved by washing the peritoneal cavity with Phosphate Buffered Saline (PBS).

Hepatocyte isolation

For in vitro studies hepatocytes were isolated 24 hrs after CCl₄ using the modified 2-step perfusion method as previously described [30,31]. Hepatocytes from animals without liver injury were isolated by the same method. Cells were cultured in DMEM medium, supplemented with 10% FBS and 1% P/S (Invitrogen, USA) in tissue culture flasks for 24 hrs prior to co-culture experiments.

in vitro co-culture

Encapsulated BMMC (1x10⁵) were co-cultured with hepatocytes (3x10⁶) from rats with CCl₄-induced liver injury (cCCl₄ group, n=3) and controls (cControl group, n=3). In addition to the capsules, both cell types were separated by a semipermeable transwell membrane (70 µm filter; BD, USA) placing the hepatocytes in the lower chamber and the BMMC in the upper chamber. The cells were co-cultured in DMEM medium supplemented with 10% FBS and 1% P/S for 6 hrs.

To evaluate the liver injury, cultured hepatocytes were stained with Oil Red O (ORO, MP Biomedicals, USA) to identify intracellular lipid droplets (ILD). Hepatocytes were fixed with formalin. After washing, 1 mL of 100% PEG (Ineos, Germany) was added for 2 min, and ORO/PEG (0.5%) was added for 10 min. After that, cells were rinsed in 60% PEG for 1 min and washed. Hepatocytes were counter-stained with hematoxilin and observed under light microscope immediately [32].

Urea quantification

After co-culture, encapsulated BMMC were removed from hepatocyte contact, washed with PBS and placed in fresh medium for 2 h. The medium was collected and urea was quantified by Quanti ChromTM Urea Assay Kit (DIUR-500, BioAssay Systems, USA) according to the manufacturer’s instructions.

Reverse transcription-polymerase chain reaction (RT-PCR)

Encapsulated BMMC recovered from in vivo and in vitro experiments were dissolved using a 100-mM sodium citrate (Labsynth, Brazil) solution prior to RNA extraction. Total RNA was extracted using the RNeasy RNA isolation kit (Qiagen, Germany), and Reverse Transcription (RT) was carried out with 1 µg of RNA. Complementary DNA (cDNA) was synthesized using Superscript II RNA-reverse transcriptase (Invitrogen, USA).

For the RT-PCR, 2 µL cDNA-templates were mixed with 5 µL PCR-buffer, 1.5 mM µl MgCl₂, 10 mMol dNTPs, 20 pmol of each primer, and 2U Taq DNA polymerase in a total volume of 50 µL. All reagents were from Invitrogen (USA). PCR was carried out using primers and conditions showed in Table 1. Samples were analyzed on 1.5% agarose gels stained with ethidium bromide.

Table 1: Primers and amplification conditions used to analyze gene expression of BMMC

Statistical analysis

Statistical comparison of urea levels of the cControl and cCCl₄ groups was carried out using Student’s T-test, using the SPSS v.18. The level accepted for significance was p<0.05.

in vivo experiments

In order to study BMMC differentiation, encapsulated cells were implanted in the peritoneal cavity of Wistar rats 24 hrs after CCl₄ or olive oil administration. In animals from tCCl₄ group, liver injury was confirmed at the time of surgery by the characteristic nutmeg aspect of the liver. Animals from tControl group didn’t present this feature throughout the study. Animals were sacrificed in CO₂ chamber after 6, 24 or 48 hrs and capsules were retrieved from the peritoneum. Capsules were found freely disseminated in the peritoneal cavity and it was not observed liver tissue from the recipient attached to the capsules after explantation.

Gene expression pattern was markedly different in BMMC from tCCl₄ retrieved after 48 hrs, showing the expression of hepatocyte markers such as Albumin and Cytokeratin 18, but negative for Alfa-fetoprotein (Figure 2). On the other hand, BMMC retrieved at 6 or 24 hrs after injections were negative for all markers but Gapdh. Interestingly, BMMC from tControl group were also negative for hepatocyte markers at all time points, except for the internal control (Gapdh), showing an expression pattern similar to that of naïve BMMC.

Figure 2: BMMC expression of hepatocyte markers 48 hrs after implantation in animals with CCl₄-induced acute liver injury. RTPCR analysis of gene expression in encapsulated BMMC from tControl (lane 1) or tCCl₄ (lane 3) retrieved 48 hrs after implantation. Positive control (hepatocytes, lane 2)

in vitro experiments

Since encapsulated BMMC showed the expression of hepatocyte markers 48 hrs after implantation in animals with CCl₄-induced acute liver injury, we tried to replicate the same model in vitro in order to determine when this reprogramming takes place.

As a way to assess hepatocyte damage in this model, cells isolated from animals with or without liver injury were stained with oil red. In CCl₄ treated animals, esteatose could be observed with several intracellular lipid droplets, that were absent in control animals. Furthermore, we observed the macroscopic aspect of nutmeg liver by the time of hepatocyte isolation.

Encapsulated BMMC that were in co-culture with hepatocytes from CCl₄ treated animals for only six hrs showed Albumin expression, whereas those in contact with control hepatocytes were negative (Figure 3). No expression of Ck18 or Afp was detected.

Figure 3: RT-PCR analysis of Alb gene expression in BMMC after 6 hour co-culture. Hepatocyte (positive control, lane 1). Encapsulated BMMC from cCCl₄ group (lane 2). Encapsulated BMMC from cControl group (lane 3). Gapdh: internal control

In order to determine if this reprogramming also had functional repercussion in BMMC we tested urea production by these cells. After co-culture with hepatocytes, encapsulated BMMC were kept for other 2 hrs in culture, with fresh medium. Urea production was greatly increased in BMMC from cCCl₄ group (Figure 4) as compared to cControl group. However, it was lower than the amount produced by hepatocytes in culture (5.85 mg/dL).

Figure 4: Quantification of urea production in the culture medium from encapsulated BMMC. Urea production was increased in BMMC from cCCl₄ group but not in cControl. Medium from hepatocyte culture was used as reference value. * p =0.033 compared to cControl group (Student’s T-test)

The ability of BMMC to differentiate into hepatocyte-like cells has been shown both in vivo and in vitro by different groups. In the present study we have shown that BMMC are able to express hepatocyte-specific genes after 48 hrs of transplantation in CCl₄ treated animals. in vitro, these cells express Albumin and produce urea after only 6 hrs of co-culture with injured hepatocytes.

Several studies have shown the ability of BMMC to differentiate into hepatocyte-like in vivo. Most of these studies show the expression of Alb and Afp [12-15,17,33], but also Ck8 and Ck18 [15]. As in the previous studies, this differentiation occurs only when BMMC are transplanted into injured animals, suggesting that liver damage triggers the release of substances that modulate BMMC gene expression. However, unlike ours, other groups describe differentiation after 13 [34] and 60 days [12,13].

There is still debate in the literature if these cells convert into hepatocyte-like by means of fusion or differentiation [18]. Our results favor the differentiation hypothesis, in accordance with other study that used microencapsulated cells [15]. Liu and Chang [15] have shown that BMMC encapsulated with the APA (Alginate-Polylysine-Alginate) method improve survival in a 90% partial hepatectomy rat model of acute liver failure. In addition, they found that some BMMC express ALB, AFP, CK 8, CK 18 and were able to store glycogen, after 2 weeks of transplantation. In our study, we used similar alginate microcapsules to isolate BMMC and we observed differentiation signs in a shorter period of time (48 h). It is important to highlight that we assessed differentiation by gene expression analysis, whereas the previous study by Liu and Chang used immunocytochemistry. It was not possible to prove the expression of ALB, AFP nor CK18 by immunocytochemistry due to difficulties in capsule’s histology. It cannot be ruled out, however, that fusion mechanisms do exist or play a role in cell-mediated tissue recovery after injury. Yet, in this work alginate microcapsules prevented any type of fusion between BMMC and hepatocytes, although allowing for the interchange of soluble compounds between these two cell types.

in vitro differentiation is usually obtained by the use of growth factors, especially Hepatocyte Growth Factor (HGF), although quite a number of differences exist between protocols [19,23-25]. Some authors have shown differentiation also with hepatocyte conditioned-medium [21] or serum from patients with hepatitis B virus-associated liver cirrhosis [35], co-culture with hepatocytes [36-38] or liver fragments [39] from healthy or injured animals. However, unlike the present study, the above mentioned groups use MSC and not BMMC. Even though MSC have a well known differentiation capacity, their isolation, culture and expansion are time consuming processes and changes in gene expression at early times like this are not reported by studies with MSC, which usually differentiate after 3 to 40 days in culture [22,36]. On the other hand, BMMC are readily available, and does not need a culture step prior to administration.

Our results showed that BMMC co-cultured with injured hepatocytes for six hrs showed expression of Albumin and urea production, both characteristics of hepatocyte activity. It has been reported that kidney tubular epithelium also produces urea, while extraembryonic cells express albumin, however, only hepatocytes can do both [40,41]. In contrast, when BMMC were co-cultured with hepatocytes isolated from healthy animals, no Albumin expression or urea production was detected after 6 hrs. Other reports have shown differentiation of MSC even after co-culture with healthy hepatocytes [36]. These differences may be reconciled by intrinsic characteristics of the cell types used in this study or by the short period of time analyzed. On the one hand, BMMC are composed of phenotypically and functionally different cell populations and we have preliminary data indicating that is the non-adherent fraction of bone marrow mononuclear cells that differentiates. However, further characterization of these cells is still pending. On the other hand, it is possible that injured hepatocytes secrete paracrine factors that lead to BMMC reprogramming sooner than healthy hepatocytes [42,43].

Interestingly, in vivo differentiation occurred only after 48 hrs. It is possible that in vitro hepatocytes secrete factors at higher concentrations, thus inducing differentiation in shorter times.

In summary, this work shows that BMMC are able to differentiate into hepatocyte-like cells in a short period of time both in vivo as in vitro. This differentiation occurs without fusion and is triggered by factors present only in injured liver. The identification of such factors and of the cells in the mononuclear fraction that respond to them is under further investigation.

This work was supported by: Research Incentive Fund of Hospital de Clínicas de Porto Alegre (FIPE-HCPA No.08-695); The National Council for Scientific and Technological Development of Brazil (CNPq), Fund for Research Support of the State of Rio Grande do Sul (FAPERGS) and Program for Support of Centers of Excellence (PRONEX – FAPERGS/CNPq n. 008/2009).