The effect of hepatocyte growth factor on intestinal adaption in an experimental model of short bowel syndrome

Nowadays, the standard therapy for patients with short bowel syndrome is parenteral nutrition (PN). Various growth factors have been tested to achieve weaning from prolonged PN administration. We evaluated the effect of hepatocyte growth factor (HGF) on structural intestinal adaptation and cell proliferation in a rat model of SBS. Thirty Sprague–Dawley rats were divided into three groups; group A rats (sham) underwent bowel transection, group B rats underwent a 75% bowel resection, and group C rats underwent the same procedure but were treated postoperatively with HGF. Histopathologic parameters of intestinal adaptation were determined, while microarray and rt-PCR analyses of ileal RNA were also performed. Treatment with HGF resulted in significant increase in body weight, while the jejunal and ileal villus height and crypt depth were increased in HGF rats (36%, p < 0.05 and 27%, p < 0.05 respectively). Enterocyte proliferation was also significantly increased in HGF rats (21% p < 0.05). Microarray and quantitative rt-PCR analyses showed that the genes hgfac, rac 1, cdc42, and akt 1 were more than twofold up-regulated after HGF treatment. HGF emerges as a growth factor that enhances intestinal adaptation. The future use of HGF may potentially reduce the requirement for PN in SBS patients.


Introduction
Short bowel syndrome (SBS) is a rather uncommon condition of malabsorption and malnutrition resulting from the loss of intestinal absorptive surface area. In adults, SBS may be defined as a length of functional bowel less than 2 m; in infants, it is defined as the loss of 75% of the functional length which is expected for the gestational age [1]. To date, parenteral nutrition (PN) is the standard therapy for patients with SBS; it has been reported that 10-15,000 patients in Europe have PN-dependent SBS, with similar estimations in the USA [2]. However, the probability of weaning from PN 5 years is only 55% and depends on functionality of remnant small bowel [3].
An important mechanism for weaning from PN in patients with SBS is intestinal adaptation; during this phase, the intestinal absorptive surface area, as well as the peptide, glucose, and electrolyte uptake are increased. Intestinal adaptation occurs shortly after massive loss of bowel competence and its degree depends on many factors, such us the length of remaining bowel, the underlying disease, and the activation of cytokines and hormones. Over the past decades, many animal and clinical studies utilizing hormones and growth factors for augmentation of intestinal adaptation have been implemented, aiming the further increase in nutrition absorption and thus earlier weaning from PN in patients with severe SBS. Growth factor (GH), insulin-like growth factor 1 (IGF-1), and glycagon-like peptide 2 (GLP-2) have been extensively studied, with promising results [4]; in fact, GH and GLP-2 analog, teduglutide, have been approved for treatment in SBS [5]. Hepatocyte growth factor has been scarcely investigated for intestinal adaptation, although it is a multipotent cytokine, with mitogenic and morphogenic activity in many viscera [6].
Hepatocyte growth factor (HGF) and its receptor, c-MET, are expressed in various organs; the product of activation, c-MET tyrosine kinase, recruits signaling effectors, inducing multiple cell activities, such as mitogenesis, cell survival, and anti-fibrosis [7]. Recently, Sugita et al. highlighted the protective effect of HGF on intestinal mucosal atrophy induced by TPN by increasing the intestinal absorptive mucosal surface area in an animal model [8]. Moreover, it is widely accepted that activation of HGF/c-MET signaling pathway is crucial for liver regeneration, especially in cases of liver diseases, via eliciting the proliferation of hepatocytes [9]. In addition, Yano et al. conducted that HGF prevented hepatic steatosis and inflammation in an animal model of SBS, via Toll-like receptor (TLR) and Farnesoid X receptor (FXR) cascade [10]. Τhis hepatotropic role of HGF may be of high significance in patients with SBS who develop intestinal failure-associated liver disease (IFALD), one of the most devastating complications of SBS. In this experimental study, we examine the effect of HGF administration on intestinal tissue after massive small bowel resection. We examine the impact of the treatment with HGF on the parameters of intestinal adaptation and we investigate the mechanisms of these results.

Animals
The study has been approved by the Veterinary Department of the Region of Attica under the code 441/22-01-2016 and the local ethics committee under the code 1281/1-2-16. Animal care complied with the Guide for the Care and Use of Laboratory Animals. The Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines were consulted throughout the conduct of this study [11]. Thirty Sprague-Dawley, non-genetic modified female rats, obtained from Hellenic Pasteur Institute, weighing 200-250 g were kept in stainless steel cages in a controlled environment (22 °C with a 12-h light/dark cycle) for at least 7 days. The animals had free access to water and standard rat chow. The rats were divided in 3 groups: group A (sham), consisted of 10 rats who underwent small bowel transection and reanastomosis; group B (MSBR), consisted of 10 rats who underwent 75% small bowel resection followed by end-to-end jejunoileal anastomosis; group C (MSBR -HGF), consisted of 10 rats who underwent the same procedure, but were treated postoperatively with HGF. The number of animals per group was dedicated using IBM SPSS Sample Power 3.0.

Experimental protocol
All surgical procedures were performed after appropriate animal preparation and under sterile conditions. The animals received general anesthesia with ketamine (50 mg/kg) and xylazine (14 mg/kg) administered intramuscularly and the abdomen was opened with a midline incision. Group A rats underwent small bowel transection with reanastomosis, 7 cm proximal to ileocecal valve. In group B and group C animals, a 75% resection of the small intestine was performed, preserving the vascular arcade and leaving 5 cm of jejunum distal to Treitz ligament and 7 cm of ileum proximal to ileocecal valve. The intestinal continuity was restored by an end-to-end single layer anastomosis using interrupted 5/0 monofilament silk sutures (SMI, Belgium). Finally, the abdominal cavity was closed in two layers with interrupted 3/0 Vicryl sutures. Immediately after operation, the rats were resuscitated with an intraperitoneal injection of 3 ml 0.9% saline solution.
Postoperatively, the animals had access only to water for the first 2 days following operation. From the 3rd postoperative day, rats were fed with standard (10% kcal fat) diet and water ad libitum. Starting from the 4th postoperative day, the animals of group C were treated with HGF administered intraperitoneally at a dose of 250 μg/kg/day; this dose was consistent with the optimal dosage of HGF administered intraperitoneally, determined in previous studies for enhancement of intestinal adaptation [12,13]. General condition, body weight, and food intake were monitored on a daily basis. Neither complications nor deaths occurred during the postoperative period. At 15 th postoperative day, the animals were anesthetized with ketamine (50 mg/kg) and afterward were sacrificed applying open pneumothorax. After assuring that no peritonitis was developed, the remnant small bowel was measured, using a suture extending from Treitz ligament to ileocecal valve. Then, the entire small bowel was resected, and tissue sections were obtained transversely at least at one centimeter away from the anastomosis at the jejunum and the ileum. The rest bowel was cut longitudinally, rinsed with saline, weighted, and harvested. Bowel weight was expressed as weight per unit of bowel length per 100 g body weight (mg/cm/100 g) to reduce bias.

Morphometry
For the histopathological examination, we used the samples from both jejunum and ileum which were obtained as described above. The samples were fixed for 24 h in 10% formalin solution, and then were dehydrated with ethanol and xylene and embedded in paraffin cubes. Subsequently, 3.5 µm-thick sections from paraffin cubes were cut using a microtome and stained with hematoxylin-eosin staining. The sections were examined microscopically using an light microscope (Nikon Eclipse 80i, Nikon Corp, Tokyo) equipped with a digital camera (Nikon Corp.) under 10 × 2 original magnification. The samples were then evaluated by an experienced pathologist and they were considered well oriented when the muscularis propria was evident, the villi were developed within the lumen, and the crypts were cut longitudinally. The images were then furtherly magnified (10 × 4) and analyzed with the Image Pro Plus 5 software (Media Cybernetics, Baltimore, MD, USA). From each histological section, an experienced pathologist identified ten villi well-distended from base to tip and ten crypts. Then, the distance from each base to tip for villi and from each base to depth for crypts was measured in μm and the mean height or depth was recorded.

Immunohistochemistry
The rate of enterocyte proliferation was assessed with Ki-67 antibody (MIB-1, Ki-67 concentrated and prediluted rabbit monoclonal antibody; dilution 1:50; Biocare Medical, Walnut Creek, CA). From the prepared paraffin cubes, additional 3.5 µm-thick sections were cut and placed on poly-Llysine coated slides. The sections were then be deparaffined, hydrated, and placed in Tris/EDTA buffer, pH 9. After being rinsed with water for 5 min, the sections were incubated in 3% hydrogen peroxide for 10 min in a dark room, to inhibit the endogenous peroxidase activity and then were rinsed again with water and TBST for 5 min. The sections were thereafter incubated with Ki67 primary monoclonal antibody (MIB-1), rinsed with TBST three times and then were incubated with rabbit HRP-labeled polymer (Envision + System-HRP, DakoCytomation, Glostrup, Denmark) for 45 min. After another immersion in TBST for 5 min, the placement of DAB chromogen for 5 min and another rinsing with TBST followed. Hematoxylin was used to anneal the sections for 40 s, which thereafter was rinsed with plenty of water for 5 min. Finally, the sections were dehydrated with graded alcohol solutions, immersed in xylene, and covered with a gasket using GLC fixing agent. All samples were successfully stained with this process. The proliferation index (PI) was defined as the number of proliferating crypt cells (stained positively for Ki-67) per 10 crypts.
The apoptosis rate was determined with monoclonal antibody caspase-3 (caspase-3 cleaved concentrated polyclonal antibody; dilution 1:100; Biocare Medical,Walnut Creek, CA). Additional 3.5 µm-thick sections were cut and, using a combination of the streptovidin-biotin-peroxidase method and microwave antigen retrieval on formalinfixed, paraffin-embedded tissues according to the manufacturer's protocols were further analyzed. All samples were successfully stained with this process. Based on the above immunohistochemical staining, the apoptotic index (AI) was defined as the number of apoptotic cells per 10 villi.

Microarray analysis
From each rat, an additional part from ileum was rinsed with saline, and using a glass slide, the intestinal mucosa was collected in a RNA stabilization reagent (RNAlater® Thermo Fisher Scientific, Waltham, MA, USA) and the sample was immediately frozen. The RNA isolation was performed according to the TRI Reagent® protocol (Sigma-Aldrich, St. Louis, MO, USA) and the concentration of RNA in the sample was determined by spectrophotometer. Then, part from the total RNA was amplified into cRNA and biotinylated by in vitro transcription using the TargetAmp Nano-g Biotin-aRNA labeling kit for the Illumina system (Picentre Biotechnologies, San Diego, CA, USA) according to the manufacturer's protocol, with 200 ng of total RNA as input material. Biotinylated cRNAs was purified, fragmented, and subsequently hybridized to an Illumina RatRef-12 Expression BeadChip V1 (Illumina Inc., San Diego, CA, USA) for genome-wide expression analysis containing 21,910 probes selected primarily from the NCBI RefSeq database (Release 16) according to the Direct Hybridization assay. The hybridized chip capable of querying 12 samples in parallel was stained with streptavidin-Cy3 (Amersham™, Sigma-Aldrich, St. Louis, MO, USA) and scanned with an Illumina BeadArray 500GX Reader. The scanned images were imported into Genom-eStudio (Illumina Inc.) for extraction and quality control, generating an output file for statistical analysis.

Real-time PCR
The remnant purified RNA was redissolved and DNase (Thermo Fisher Scientific, Waltham, MA, USA) was used to remove DNA traces. The RNA was transcribed into cDNA according to the PrimeScript RT Reagent kit gDNA eraser protocol (Takara, Kioto, Japan). The cDNA, along with genes which were up-regulated or down-regulated with more than twofold change, were amplified by Real-Time Polymerase Chain Reaction (PCR) according to the Quantitect GRE protocol PCR kit (Qiagen, Hilden, NRW, Germany) on Applied Biosystems 7500 fast. GAPDH was used as a housekeeping gene.

Statistical analysis
All data are expressed as means ± standard error (SE). We ran a power analysis to define the number of each study groups using IBM Sample Power 3.0, as mentioned above. Based on previous similar studies [14], we estimated that enrolling 30 animals (10 animals per study group) would allow a detection of a 15% higher enterocyte proliferation rate in animals underwent massive small bowel resection than in sham animals with a two-sided significance level of 0.05 and a power of at least 90%. Moreover, we estimated that assigning 10 animals in HGF-MSBR group would allow a detection of 25% higher enterocyte proliferation rate than in animals underwent massive small bowel resection, with a two-sided significance level of 0.05 and a power of at least 90%. The one-way ANOVA test was used for comparison, followed by Tukey's test for pairwise comparison, while the Prism software (GraphPad Software, San Diego, CA) was used for statistical analysis. The statistical significance was defined as p < 0.05. Microarray data were analyzed using the Partek Genomics Suite software program. To evaluate the magnitude of differential gene expression, the displacement of each detected transcript's mean expression value was measured between the two groups. A standard regression analysis was performed on the MSBR and MSBR-HGF groups to test whether the mean transcription level differed from that of the Sham group.

Gross morphology
The fluctuation of body weight after surgery in each study group is presented in Fig. 1. Sham (group A) animals maintained their body weight during the first 4 postoperative days, with gradual increase thereafter. Group B (MSBR) animals showed rapid weight loss during the first 4 postoperative days, with gradual increase thereafter. Rats treated with HGF demonstrated an even faster increased weight gain rate compared not only to sham (group A) rats (21.8% vs 15.5% at day 15th, p < 0.01) but also to group B (MSBR) rats (21.8% vs 17.5% at day 15th, p < 0.05).The intestinal  weight at the 15th postoperative day differed also significantly between three study groups (Table 1). Interestingly, HGF-treated rats showed significant increase in intestinal weight, not only compared to sham (group A) rats (144 ± 12 vs 49 ± 4, mg/cm/100 g, p < 0.001), but also compared to group B (MSBR) rats (144 ± 12 vs 94 ± 7 mg/cm/100 g, p < 0.05).

Intestinal adaptation
The histopathological analysis showed that massive small bowel resection was correlated with enhanced intestinal adaptation (

Microarray analysis
From the total amount of 21,910 probes, 165 genes were differentially expressed in animals of group B (MSBR) compared to sham animals. From these, 91 genes were up-regulated and 74 genes were down-regulated, while 10 genes were more than twofold regulated. In HGF-treated rats (group C), a total of 212 genes were differentially expressed, with 127 up-regulated genes and 85 downregulated genes; 16 genes were more than twofold regulated. Functional clustering analysis revealed increased expression in signaling pathways involved in aminoacid (tryptophan, glycine, serine, and threonine) and fatty acid metabolism, as well as propanoate, ascorbate, and aldarate metabolism and biosynthesis of steroids. Next, we investigated 22 genes related to HGF/c-MET signaling pathway (Table 3). In animals of group B (MSBR), 7 genes were up-regulated with a relative change in gene expression level of 10% or more, 12 genes remained practically unchanged, and 3 genes were down-regulated.
In HGF-treated animals (group C), 16 genes were upregulated, 5 genes remained unchanged, and 1 gene was down-regulated.

Real-time PCR
Of the 22 genes associated with the HGF/c-MET pathway, 5 genes were more than twofold up-regulated (p < 0.05) after treatment with HGF, according to microarray analysis. These genes were: hgfac (Hepatocyte Growth Factor Activator), rac 1 (Ras-related C3 botulinum toxin substrate 1), cdc42 (Cell Division Cycle 42 gene), stat-3 [Signal Transducer And Activator Of Transcription 3 (Acute-Phase Response Factor], and akt 1 (serine-threonine protein kinase AKT1); these genes were further investigated with real-time quantitative PCR. The results confirmed the fold-changes compared to sham (group A) animals (p < 0.05). Interesting is though, that all genes except stat-3 were statistically significantly more up-regulated in the animals treated with HGF (group C) compared to animals which had undergone massive small bowel resection (group B) (p < 0.05) ( Table 4). All study results can be found in our institutional Databank and are also available at Figshare Data Repository.

Discussion
In this study, we investigated the possible effect of hepatocyte growth factor (HGF) on intestinal adaptation after massive small bowel resection. In the literature, it is well reported that treatment with hormones and growth factors, such as GLP-2, IGF, and EGF, can augment the intestinal adaptation process; yet, little has been known about HGF. The experimental model that we used mimics the massive small bowel loss which is prominent in short bowel  Our results indicate that HGF has a positive effect on all parameters of intestinal adaptation. Moreover, admission of HGF provokes the encoding and secretion of specific proteins which play important roles in the intestinal adaptation process.
Hepatocyte growth factor, also known as scatter factor, was first described in 1984 in two independent studies [17,18] and its cDNA was first cloned in 1989 [19]. It is secreted from non-parenchymal cells as an inactive form (pro-HGF), which can be activated by proteolysis [20]. The proteolysis is mediated by various proteases, including HGF activator (HGFA), coagulation factors XII and XI, kallikrein, hepsin, and plasminogen activators. In its active form is a heterodimer composed of a 69-kD α-chain and a 34-kD β-chain [21]. The α-chain contains an N-terminal hairpin domain and four subsequent kringle domains, and the β-chain contains a domain similar to serine proteases, though without proteolytic function. The kringle domain is also similar to this found in serine proteases for thrombolysis; thus, HGF has a 38% amino acid homology to plasminogen [21]. HGF receptor, tyrosine kinase c-Met, is the product of the proto-oncogene c-met and is expressed in the parenchymal cells of many viscera, such as liver, intestine, kidney, and bone marrow [22]. In its active form is an α/β disulphide-linked heterodimer with a 50-kD α-chain and a 145-kD β-chain. The extracellular part has three domain types; a large semaphoring (Sema) domain, the PSI domain, and the IPT domain which connects the PSI domain with the transmembrane helix. Intracellularly, the c-MET receptor contains a tyrosine kinase catalytic domain flanked by distinctive juxtamembrane and carboxy-terminal sequences [7]. The binding of HGF to c-Met induces the activation of this tyrosine kinase, which recruits signaling effectors eliciting paths of cell proliferation, cell survival, cell differentiation, and morphogenesis. Shortly after a massive loss of intestinal length, the intestine undergoes an adaptive response to increase the functionality of the remnant mass. In humans, intestinal adaptation begins 16 h after loss of intestinal mass, and is mainly characterized by increased cellular proliferation [23]. The second phase of intestinal adaptation starts 1-2 weeks after intestinal loss and is characterized by cellular differentiation resulting in increase of absorptive area, and glucose and electrolyte uptake [24]; this phase will take place for 4-6 weeks. At the beginning, the mesenchymal cells, which are located under the epithelium in the crypts, are proliferating and condensing there. As these cells start to differentiate, they migrate toward the central lumen to form the villi; therefore, each villus contains epithelial cells from the adjacent crypts [25]. Differentiation and migration of these cells are mediated by various signaling pathways, such as Wnt/β-catenin [26], Notch [27], and Hedgehog [28]. This hyperplasia results in increased absorptive competence per unit of intestinal length, as the new epithelial cells contribute to amino acid, glucose, and electrolyte uptake; moreover, the villi are elongated and the crypts are deepened; thus, the absorptive surface area is increased. At the same time, the number of goblet cells in the villi is increased, probably through switch of some enterocytes [25]. Goblet cells produce epithelial-protective mucins that are involved in intestinal wound healing [29]. In the same fashion, Paneth cells' absolute number increases in the crypts, contributing to intestinal adaptation process by producing stimulating factors [23].
The extend of intestinal adaptation ranges between cases and it is believed that it is proportional to the amount of intestine loss [30]. Therefore, it comes as no surprise that massive small bowel resection enhanced almost all parameters of intestinal adaptation. These findings are consistent with the findings from our previous study [14] and from other similar studies [15,31] and prove the efficacy of our experimental model. Interestingly, the chances in morphometry were more prominent in ileal samples; this is consistent with previous findings, suggesting that distal ileum adapts faster and demonstrates greater morphological changes [32,33]. The proliferation and apoptosis rate were also higher in ileum compared to jejunum, finding that is consistent with previous studies, which have shown that proliferation and apoptosis are relatively more evident in ileum than in jejunum [34,35].
A finding of special interest was that the massive intestinal resection triggered the up-regulation of many genes belonging to the HGF/c-MET system, suggesting the participation of this pathway in the stages of intestinal adaptation [36]. From them, of specific interest was the up-regulation of stat-3 gene. STAT-3 is a pleiotropic transcription factor that regulates cell stress response and promotes wound healing [37]. STAT-3 is phosphorylated by receptor-associated Janus kinases (JAK) in response to cytokines and growth factors, including HGF, and translocate to the cell nucleus where they act as transcription activators [38].
Since HGF/c-MET signaling activates a wide range of different cellular pathways, we hypothesized that HGF promotes intestinal tissue regrowth after massive bowel resection. The animals treated with HGF had significantly greater body weight gain rate compared not only to sham animals but also to animals underwent massive small bowel resection. Additionally, HGF-treated rats had significantly higher intestinal weight compared to group A and group B rats. Villus height and crypt depth were also significantly increased compared to untreated animals. Moreover, cellular proliferation was enhanced, while apoptosis was attenuated in HGF-treated animals resulting in greater augmentation of intestinal hyperplasia. This augmentation can be explained by the regulation of genes associated with cell proliferation, cell survival, cell differentiation, and migration. Microarray analysis showed that a total of 212 genes were differentially regulated after administration of HGF, while real-time PCR showed a twofold increase of hgfac, rac1, CDC42, akt1, and stat-3. Of particular interest is though the twofold increase of rac 1 and cdc 42. Both CDC42 and Rac 1 are involved in regulation of the signaling pathways that control diverse cellular functions including cell morphology, migration, cell-cell adhesion, endocytosis, and cell cycle progression; thus, they promote cell proliferation. Furthermore, CDC42 and Rac1 are key signaling switches to actin cytoskeleton reorganization [39] that ensure coordinated control of different cellular activities such as gene transcription and adhesion through their interaction with multiple target proteins [40].
The up-regulation of akt 1 after treatment with HGF is also a remarkable finding. The PI3-kinase/Akt signaling pathway has a significant anti-apoptotic role; akt phosphorylates the IκB kinase, which results in IκB degradation and allows NF-kB to enter the nucleus and activate transcription of anti-apoptotic genes [41]. In that way, cell apoptosis is attenuated. Another possible mechanism of the anti-apoptotic effect after HGF administration is the regulation of MAPK pathway. HGF/c-MET elicits the activity of RAS and via the Gab1/Grb2/SOS cascade, and the MAPK effector kinase and ultimately MAPK are activated. MAPK pathway promotes cell scattering and tubulogenesis [42], but many studies have shown that is also involved in an anti-apoptotic signaling elicited by HGF [37].
The idea of using growth factors for augmentation of intestinal adaptation is not new; several studies, both experimental and clinical, have delineate the role of glycagon-like peptide 2 (GLP-2) in intestinal adaptation [4]. In fact, teduglutide, the GLP-2 analog, has been approved for treatment in patients with SBS [5]. Growth factor (GH), insulin-like growth factor 1 (IGF-1) [5], epidermal growth factor (EGF) [42], and glycagon-like peptide 1 (GLP-1) [4] have been also extensively studied, with decent results in terms of intestinal adaptation. On the contrary, HGF has not been examined in that extent. An interesting study from Kato et al. conducted that treatment with HGF increased galactose and glycine absorption, as well as overall mucosal mass in rats after massive small bowel resection [43]. From the same institution, Katz et al. reported similar results with significant increase of morphological parameters of intestinal adaptation after HGF treatment in an experimental model of short bowel syndrome [44]. Recently, Satoshi Ieiri institution has also examined the role of HGF in mucosal atrophy in animal models. Sugita et al. [8] used TPN alone for inducing mucosal atrophy, and the managed to show that the jejunal villus height and the crypt cell proliferation rate in animals treated with HGF were significantly higher compared to controls; moreover, animals which received HGF in high doses had significantly higher intestinal absorptive mucosal surface compare to controls. In our study, the effect of HGF in intestinal adaptation was correlated with specific genetic pathways, providing more information about the HGF mechanisms on intestinal tissue.
Although the role of HGF in intestinal adaptation has not been extensively studied, its role in other viscera and especially in liver has been decently delineated. Kinoshita et al. reported that HGF mRNA levels are increased in response of extended hepatectomy in rats [45]. HGF acts as an antinecrotic factor, which inhibits cholestatic hepatitis and also acts as an antifibrotic factor which inhibits the hepatic fibrosis process by stimulating the hepatocyte proliferation [46]. This hepatotropic role of HGF may be of high significance in patients with short bowel syndrome who develop IFALD. Yano et al. conducted that treatment with HGF in rats which underwent small bowel resection resulted in the prevention of hepatic steatosis and inflammation [10]. In particular, treatment with HGF resulted in overexpression of FXR, a receptor whose overexpression in liver alleviates liver damage [47]. Yano et al. also showed that TLR4 expression was also lower in animals treated with HGF; this finding is of high significance, as TLR4 signaling is crucial for developing IFALD [48]. Therefore, utilization of HGF in patients with short bowel syndrome may be greatly beneficial, as not only enhances intestinal adaptation, but also has a hepatoprotective effect.
Regarding the safety of HGF gene therapy, even from the preliminary studies examined HGF, it is known that a brachytherapy with HGF is considered safe, as HGF is rapidly cleared from the circulation due to its bond with heparin sulfate [49]. In our study, we did not observe any clinical or histopathological adverse event. Studies in rats conducted that prolonged treatment with HGF in high doses may cause renal injury; yet, a brachytherapy with HGF does not seem to affect renal sufficiency [50]. Since it is a growth factor which enhances cell proliferation and alleviates cell apoptosis, HGF treatment may have potential deleterious effects, such as hepatic or intestinal malignancy; yet, they can be avoided, as the HGF treatment would last a limited time. The cessation of gene expression beyond this time point can be considered to constitute an inherent safety feature of HGF gene transfer. All in all, no major clinical study with long-term follow-up questioning the long-term safety of HGF therapy has been undertaken, and more large, randomized trials are needed to establish not only the efficacy, but also the safety of HGF therapy.
There are a number of limitations of this study. First of all, we administered the HGF intraperitoneally; although it is a well-established method for animal models, it is inferior to intravenous administration of HGF. Intravenous administration of HGF via a central venous catheter allows same or even more prominent results with less dose, thus less adverse effects. Another limitation of the study was the postoperative period of 15 days. A longer postoperative course of, e.g., 21 or 30 days would allow to examine better the effect of HGF in intestinal parameters over a period of time. Finally, a limitation of this study, is that the genes identified by microarray and PCR analyses do not necessarily have physiologic importance. Moreover, the magnitude of the fold change is not necessarily associated with the effect that a gene may have on the animal. Nevertheless, further studies are needed to better understand the role of HGF in intestinal adaptation.
In conclusion, treatment with HGF enhances intestinal adaptation after massive small bowel resection. Cell proliferation is augmented and, therefore, the intestinal absorption surface area is lengthened. Our analysis showed that the activation of HGF/c-MET pathway results in upregulation of specific genes which are related with cell proliferation and differentiation.