TIMP-1 As a Promising Therapeutic Indicator for the Intestinal Inflammation and Fibrosis.


 Background: Intestinal fibrosis is a major complication of inflammatory bowel disease (IBD), which can be prevented with anti-inflammatory therapies. However, inflammatory biomarkers do not necessarily reflect the progression of fibrosis, and hence, an indicator that reflects the therapeutic effects of available anti-inflammatory drugs on fibrosis is needed. This study shows that tissue inhibitor of metalloproteinase-1 (TIMP-1) is a potential therapeutic indicator for ongoing intestinal fibrosis in IBD patients.Methods: TIMP-1 mRNA levels in human myofibroblasts were measured, and TIMP-1 expression in the intestine of 10 Crohn’s disease (CD) patients was examined by immunohistochemistry. Colitis was induced in C57BL/6 mice by administration of 1.5% dextran sodium sulfate for 7 days; then, the mice were treated with anti-TNF-α antibody or phosphate-buffered saline for 14 days. Intestinal fibrosis was evaluated by Masson’s trichrome staining, and TIMP-1 mRNA levels in the intestinal tissue and plasma TIMP-1 levels were measured.Results: TIMP-1 expression in human fibroblasts increased after differentiation into myofibroblasts. TIMP-1 expression in CD intestinal tissue was significantly higher in the inflammation area than in the fibrotic or normal areas. In a murine colitis mode in which TIMP-1 mRNA levels were increased, anti-TNF-α antibody administration significantly attenuated inflammation and fibrosis assessed by a reduction in TIMP-1 mRNA levels in intestinal tissue and plasma TIMP-1 levels.Conclusions: The results in this study show that TIMP-1 is a potential new biomarker that could be used to assess the effectivity of different therapeutic strategies for intestinal inflammation and fibrosis in IBD patients.

. In iximab has shown to have a preventive effect on the recurrence of CD after an endoscopic procedure, including preventing the luminal narrowing 18-24 months after surgery for CD patients [8,9]. However, these compounds do not reduce the incidence of intestinal strictures in CD [2,[10][11][12][13], and therefore, developing new therapies that prevent intestinal brosis and reduce the incidence of intestinal strictures is still necessary. When both in ammatory biomarkers and clinical symptoms are considered to manage the in ammation alterations in CD patients, outcomes are better than when decisions are only symptom-driven [14]. Similarly, treatment decisions for preventing intestinal brosis should also be based on objective indicators like biomarkers levels. However, existing in ammatory biomarkers such as Creactive protein, erythrocyte sedimentation, serum amyloid A, immunological fecal occult blood test, and fecal calprotectin are not suitable biomarkers to be used in this context. As there is a complex interface between in ammation and brosis, including cytokine networks; the anti-in ammatory effect of a certain therapy and the in ammatory biomarkers pro le do not necessarily re ect the therapeutic effect of the compound. Hence, in ammatory markers cannot be used as indicators of the prevention of intestinal brosis, and it is necessary to nd new, different indicators.
Fibrosis refers to the accumulation of collagen-rich extracellular matrix (ECM) in response to tissue damage, and it is caused by repeated epithelial injury due to in ammation. The epithelial injury causes the accumulation and activation of mesenchymal cells, such as broblasts and myo broblasts, as well as the in ltration of in ammatory cells [15][16][17]. Excessive accumulation of ECM occurs due to an aberrant restoration of the tissue architecture [18]. ECM accumulation is regulated by two protein families: matrix metalloproteinases (MMPs), which degrade the ECM, and tissue inhibitors of metalloproteinases (TIMPs), which inhibit the MMPs function. Hence, the balance between brogenesis and brolysis is controlled by the reciprocal relationship between MMPs and TIMPs. During brogenesis, an excess of TIMPs over MMPs occurs, which results in the inhibition of matrix degradation [19]. TIMP-1, a member of the TIMP family, is a potent inhibitor of many MMPs [17]. Since several studies have shown TIMP-1 overexpression in pro brotic environments, such as scar tissue formation, pulmonary brosis, and diabetic nephropathy, TIMP-1 is considered to play an important role in restricting ECM proteolysis [20]. In addition, serum TIMP-1 is one of the index factors included in in the enhanced liver brosis score, which is used to evaluate liver brosis [21]. Increased expression of TIMP-1 has also been shown in brotic lesions in CD patients and in a murine model of colitis [22][23][24][25]. Since TIMP-1 is strongly associated with organ brosis and inhibits the degradation of ECM, it is possible that TIMP-1 is a suitable biomarker to predict the therapeutic effect of anti-in ammatory agents on ongoing intestinal brosis.
In the present study, we show that the expression levels of TIMP-1 re ect the prospective effect of antiin ammatory therapies that aim to prevent intestinal brosis progression using a murine colitis model. These results suggest that TIMP-1 has potential applications as an indicator of ongoing intestinal brosis in IBD.

Materials And Methods
Isolation of broblasts from human colon tissues and TGF-β treatment Human colonic broblasts were isolated from the resected colon tissue of patients undergoing endoscopic submucosal dissection for transverse colon adenoma, as described previously [26]. The broblasts were maintained in Dulbecco's modi ed essential medium and Ham' s F-12 (Wako, Osaka, Japan) supplemented with 10% heat-inactivated fetal bovine serum and 1% antibiotic-antimycotic at 37°C in an atmosphere containing 5% CO 2 . The broblasts were grown in serum-free media for 24 h and then exposed to recombinant human transforming growth factor beta (TGF-β) (10 ng/mL, R&D Systems, Minneapolis, MN, USA) for 72 h. This study was approved by the Tohoku University ethics committee (no. 2020-1-482), and informed consent was obtained from the patients prior to the study. This study was performed in accordance with the standards of the Declaration of Helsinki and the current ethical guidelines.

Patients and tissues
This study group consisted of 10 patients with established CD diagnoses who consecutively underwent surgery for intestinal strictures from 2019 to 2020 at Tohoku University Hospital. Of these patients, six were female and four were male, with ages ranging from 19 to 60 years (mean 37.6 years, Table 1).
Informed consent was obtained from all of the patients, and the Tohoku University ethics committee approved the research protocols for this study (no. 2020-4-064). This study was performed in accordance with the standards of the Declaration of Helsinki and the current ethical guidelines. Formalin-xed, para n-embedded sections were cut from the tissue to a thickness of 3 μm. Para n sections were depara nized in xylene and rehydrated with graded ethanol and distilled water. To visualize the TIMP-1 antigens, sections were heated in a microwave oven for 15 min in citrate buffer at pH 6. The nonspeci c binding sites were blocked with 1% bovine serum albumin (Invitrogen, Carlsbad, CA, USA) in phosphate-buffered saline (PBS) for 30 min. The samples were then incubated overnight at 4ºC with the primary antibody anti-TIMP-1 rabbit monoclonal antibody (Abcam, Cambridge, UK; dilution 1:300). After blocking the endogenous peroxidase activity with methanol containing 0.3% hydrogen peroxide (H 2 O 2 ), the sections were sequentially incubated with EnVision+ System-HRP (Dako, Glostrup, Denmark), and the immune complexes were visualized with 3,3-diaminobenzidine (DAB; Dojindo, Kumamoto, Japan) solution (1 mmol/L DAB, 50 mmol/L Tris-HCl buffer [pH 7.6], and 0.006% H 2 O 2 ). All sections were counterstained with hematoxylin. For the negative controls, the incubation with the primary antibody step was omitted. As expected, no detectable staining was evident. Images were captured using a BZX-800 uorescence microscope (Keyence, Osaka, Japan). To evaluate TIMP-1 expression, the percentage of TIMP-1 positive cells was calculated as follows: the number of TIMP-1 positive cells/the number of all cells in a eld of view.

Establishment of a murine colitis model
All animal experiments were approved by the Tohoku University Animal Care Committee. The study was carried out in compliance with the ARRIVE guidelines (http://www.nc3rs.org.uk/page.asp?id=1357). Nine week old male C57BL/6 mice weighing 22-25 g were purchased from SLC Japan, Inc. (Shizuoka, Japan). Mice were fed with common commercial pellet diets and ordinary tap water, and housed in an airconditioned room at a temperature of 24 °C. Intestinal brosis was induced in the models by oral administration of 1.5% dextran sodium sulfate (DSS) (MP Biomedicals, Santa Ana, CA, USA) dissolved in the drinking water. The mice were acclimatized for the rst week. At 10 weeks of age, mice were randomly divided into two groups: DSS mice treated with PBS (PBS group, n = 15) and DSS mice treated with anti-TNF-α antibody (anti-TNF-α group, n = 15). Mice received 1.5% DSS drinking water on days 0-7 and regular drinking water on days 7-21. One hundred microliters of PBS or 20 μg of mouse anti-TNF-α monoclonal antibody (Clone MP6-XT22, R&D Systems) in 100 μL PBS were injected intraperitoneally twice a week, from day 7 onwards. Mice body weight, stool consistency, and stool bleeding were monitored twice a week. Colitis severity was scored by evaluating the clinical disease activities. The disease activity index (DAI) was determined as previously described [24]: change in body weight loss (no weight loss or weight gain = 0; 5%-10% weight loss = 1; 11%-15% weight loss = 2; 16%-20% weight loss = 3; >21% weight loss = 4), stool consistency (normal and well-formed = 0; very soft and unformed = 2; watery stool = 4), and stool bleeding (normal color stool = 0, reddish color stool = 2, bloody stool = 4). The DAI score was calculated as the total of these scores, and ranged from 0 (healthy) to 12 (severe colitis). Mice were sacri ced by an overdose of sevo urane on days 7, 14, and 21 (n = 5). The distal colon was excised longitudinally. Colon tissues were stored in liquid nitrogen for quantitative real-time PCR, xed in 10% buffered formalin, and embedded in para n for histological examination.
Histological examinationof murine colitis model Murine colon tissue was xed in 10% buffered formalin, embedded in para n, and stained with Masson's trichrome stain (MTS). To evaluate brosis of the colon, area positive for MTS staining was quanti ed by measuring ve randomly non-overlapping elds from the distal colon using a BZX-800 uorescence microscope (Keyence).

Reverse transcription and quantitative real-time PCR
Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. cDNA was synthesized using the PrimeScript RT Reagent Kit (TaKaRa Bio, Shiga, Japan). Quantitative real-time PCR was performed using SYBR Premix Ex Taq II, ROX Plus (Takara Bio) according to the manufacturer's instructions using a StepOnePlus Real Time PCR System (Applied Biosystems, Foster City, CA, USA). The cycle threshold (Ct) was calculated using the comparative CT method. The relative mRNA expression was normalized to that of GAPDH. The primers used in this study are listed in Supplemental Table 1.

Enzyme-linked immunosorbent assay (ELISA)
Blood samples were collected via cardiac puncture at sacri ce. Plasma was collected by centrifugation at 1300 × g for 10 min using heparin as an anticoagulant, and plasma samples were stored at − 80 °C until analysis. Commercial IL-6 and TIMP-1 ELISA kits (Abcam) were used to measure plasma levels according to the manufacturer's instructions.

Statistical analyses
All data are presented as the mean ± standard error of the mean (SEM). Statistical analysis was performed using the Wilcoxon signed-rank test and Wilcoxon rank-sum test. Statistical signi cance was set at p < 0.05. JMP Pro software version 14 (SAS Institute Inc., Cary, NC, USA) was used for statistical analyses.

Results
TIMP-1 expression was increased after TGF-β stimulation in human colonic normal broblasts.
Human broblasts were cultured and stimulated with TGF-β to observe the differentiation of broblasts into myo broblasts. The broblasts showed no morphological changes after TGF-β stimulation (Fig. 1a). However, mRNA expression levels of brotic markers α-SMA and Col1a1 were signi cantly increased 72 h after TGF-β stimulation, which suggested that the broblasts had differentiated into myo broblasts (Fig. 1b). The mRNA level of TIMP-1 was also signi cantly increased after TGF-β stimulation, in agreement with the elevated expression levels of other brotic markers (Fig. 1b).
TIMP-1 expression in CD intestinal tissue was higher in in ammation areas than in brosis areas.
Immunohistochemical analysis using clinical samples from CD patients clari ed the features and patterns of TIMP-1 expression in the intestinal tissue. Regardless of the case, TIMP-1 positive cells were observed most frequently in the stroma; however, TIMP-1 expression levels differed according to the area in the same patient (Fig. 2a). In most cases, the in ammation area showed the highest level of TIMP-1 expression, whereas that in the brotic area was relatively low. The percentage of TIMP-1 positive cells in the in ammation area was signi cantly higher than that in the brotic area (Fig. 2b). In contrast, TIMP-1 positive cells were rarely observed in normal areas.
Anti-TNF-α antibody treatment attenuated in ammation and brosis in murine colitis model.
A DSS-induced murine colitis model was established as summarized in Fig. 3a, and subsequently, the animals were treated with anti-TNF-α antibody or PBS. Mice treated with DSS showed severe symptoms of colitis indicated by the change in the DAI score, including body weight loss (Fig. 3b, c).
Anti-TNF-α antibody treatment showed signi cant improvement effects on body weight loss on day 14 and the severity of colitis on days 11 and 14 (Fig. 3b, c). Masson's trichrome staining revealed that deposition of collagen bers in the intestinal submucosa was induced in DSS-treated mice in accordance with severe colitis, and anti-TNF-α antibody treatment reduced the thickness of submucosal brosis (Fig. 4a), indicating that anti-TNF-α antibody treatment also showed an inhibitory effect on the induced brosis. The ratio of the area containing collagen bers was signi cantly reduced by anti-TNF-α antibody treatment on day 21 (Fig. 4b).
TIMP-1 gene expression in murine colon tissue was increased after DSS administration and was reduced by anti-TNF-α antibody treatment, as with in ammatory markers.
The mRNA levels of in ammatory and brosis markers in the colon tissue of the murine colitis model changed over time. The expression of TIMP-1 in the PBS group was consistently high after DSS administration (Fig. 5), similarly to the expression levels of the representative in ammatory markers IL-1β and IL-6. Anti-TNF-α treatment signi cantly reduced gene expression levels of the in ammatory markers. In addition, the expression levels of TIMP-1 were also signi cantly reduced by anti-TNF-α treatment on day 21 (Fig. 5), in concordance with the tissue brosis evaluation by Masson's trichrome staining (Fig. 4b). In contrast, the expression level of Col1a1, another representative brosis marker, was not increased in day 7 and there was no signi cant difference in Col1A1 expression between the PBS and anti-TNF-α groups (Fig. 5).
Plasma TIMP-1 level in murine colitis model re ected effects of anti-TNF-α antibody.
Plasma levels of TIMP-1 also re ected the effects of anti-TNF-α antibody treatment in a murine colitis model. Anti-TNF-α antibody treatment drastically reduced the plasma levels of IL-6 ( Fig. 6), indicating that the systemic administration of the anti-cytokine agent has a powerful effect. In addition, anti-TNF-α treatment signi cantly reduced the plasma level of TIMP-1 on day 21 (Fig. 6). These results were consistent with the histological evaluation of intestinal brosis (Fig. 4b) and mRNA levels in colon tissue (Fig. 5).

Discussion
Advances in our understanding of the pathogenesis of brosis have led to the identi cation of many molecular factors associated with the progression of intestinal brosis, such as Col1a1 and α-SMA. Although many brotic markers that can detect intestinal brosis are known, few reports on the application of brotic markers to assess therapeutic strategies for IBD are available. Considering that current drug therapies for IBD are only able to prevent intestinal brosis as a consequence of suppressing in ammation, but are unable to reverse established brosis [28], it is necessary to identify an indicator to investigate if potential new treatments can prevent brosis. The results presented in this study suggest that TIMP-1 could be used as a clinical indicator to assess the effectivity of therapeutic strategies for intestinal brosis associated with in ammatory response in IBD.
TIMP-1 is a key molecule in brogenesis, and it has been previously used as a marker for brosis. In vitro studies using myo broblasts suggest that TIMP-1 appears is a brotic marker. In an in vitro study using human broblasts, mRNA expression of TIMP-1 increased during TGF-β-stimulated myo broblast differentiation. TGF-β is a key factor in the development of brosis and a central mediator of brogenesis by modulating broblast phenotype and function, inducing myo broblast differentiation, and promoting matrix accumulation [29,30]. Another in vitro study showed that TIMP-1 signi cantly enhanced the migratory potential of colonic myo broblasts isolated from patients with CD [31]. As these and other studies show, TIMP-1 can be considered a representative marker of intestinal brosis.
In contrast with the previously described in vivo studies, TIMP-1 appears to behave as an in ammatory marker in human and murine intestinal tissues. In this study, we found that TIMP-1 expression was signi cantly higher in the in amed area than in the brotic area of CD intestinal tissue. Moreover, in the murine colitis model, the expression levels of TIMP-1 in the colon tissue increased immediately following DSS administration, similarly to the in ammatory markers. Another study using a DSS-induced murine colitis model reported that TIMP-1 knock-out mice recovered faster after acute colitis with a lower disease activity index than wild-type mice [32]. TIMP-1 has not only an MMP inhibitory function, but also an independent immunomodulatory function [32,33]. Thus, TIMP-1 seems to be a molecule suitable as an in ammatory marker and a brotic marker, suggesting that it could be used to evaluate ongoing processes of developing intestinal brosis from in ammation. In contrast to TIMP-1, Col1A1, another representative brosis marker, had neither the characteristics of an in ammatory marker or of an ongoing brosis marker.
In the present study, the plasma TIMP-1 levels in a murine colitis model re ected the therapeutic effect of anti-TNF-α administration. Blood samples are more suitable to examine biomarkers of brosis than tissue samples due to their minimal invasiveness and availability. Additionally, plasma levels of the in ammatory marker IL-6 in the murine colitis model showed different changes than the IL-6 gene expression changes analyzed in the colon tissue, suggesting that the systemic administration of certain biological agents can have profound effects on the plasma levels of in ammatory cytokines. At the same time, it also suggests that the plasma levels of in ammatory markers do not necessarily re ect the therapeutic effect on intestinal brosis. After anti-TNF-α administration, TIMP-1 corresponded with the tissue brosis evaluated by Masson's trichrome staining, but plasma IL-6 were extremely low compared with the results of tissue brosis and IL-6 gene expression in the colon tissue. As it is possible that plasma levels of in ammatory markers undergo exacerbated changes upon the systemic administration of biological agents, in ammatory markers such as IL-6 are not suitable indicators to evaluate the therapeutic effect of those biological agents on ongoing intestinal brosis.
Further discussion is needed on using mice exposed to one cycle of DSS as a model for chronic intestinal in ammation and brosis. Although repeated DSS administration, usually during 7-10 days, is known to induce acute intestinal in ammation [34,35], it has been reported that C57BL/6 mice develop chronic colitis and intestinal brosis even after a single DSS cycle [35,36]. In contrast to C57BL/6 mice, BALB/c mice do not develop chronic colitis subsequent to the acute response, possibly as a result of a different cytokine response [35]. Hence, C57BL/6 mice treated with one cycle of DSS can be considered simple and suitable for assessing the effects of anti-TNF-α antibody administration to prevent intestinal brosis as a consequence of suppressing in ammation.
The present study has some limitations. First, the plasma levels of TIMP-1 were evaluated only in the murine colitis model, and not in IBD patients. Demonstrating that TIMP-1 changes in a similar manner in human patients is necessary to implement its application in the clinic. In addition, anti-TNF-α administration was used to treat in ammation in this animal study. Recently, new treatments for IBD have been developed, and various drugs are currently available for the treatment of IBD. To apply TIMP-1 as an indicator of the therapeutic effects of various anti-in ammatory drugs on intestinal brosis, TIMP-1 expression changes in response to these anti-in ammatory therapies needs to be examined. Therefore, further experiments are necessary to con rm the clinical utility, validity, and versatility of TIMP-1 as an effective indicator for different drugs.

Availability of data and materials
All data generated or analyzed during this study are included in this published article and its supplementary information les.
Competing interests  Statistical analysis was performed using the Wilcoxon signed rank test. *p < 0.05, **p < 0.01. score (c) of mice with or without TNF-α treatment. Data are presented as the mean ± SEM. Statistical analysis was performed using the Wilcoxon rank sum test. *p < 0.05, **p < 0.01.

Figure 4
Therapeutic effects of anti-TNF-α antibody on intestinal brosis in the murine colitis model (a) Microscopic images of Masson's trichrome staining in mice distal colon. Scale bars indicate 100 μm. (b) The graphs represent the ratios of Masson's trichrome staining-positive area to the entire area of the submucosa. The sample size of each group was n = 5. Data are presented as the mean ± SEM. Statistical analysis was performed using the Wilcoxon rank sum test. *p < 0.05.

Figure 5
Gene expression of in ammatory and brosis markers in the colon tissue of a murine colitis model mRNA levels of IL-1β, IL-6, MMP-13, TIMP-1, and Col1a1 in the distal colon tissue of a murine colitis model were measured using quantitative real-time PCR. Data were normalized to GAPDH expression levels. Data are presented as mean ± SEM. Statistical analysis was performed using the Wilcoxon rank-sum test. *p < 0.05. Figure 6