Visualization of Tumor Necrosis Factor-α Distributions Within Pressure Ulcer Tissue Using the Wound Blotting Method: A Case Report and Discussion

Wound blotting can be used to visualize the protein distribution on a wound bed through protein collection by attaching a nitrocellulose membrane to the wound surface. This study checked for consistency between the protein distributions determined by wound blotting and those determined by removal of the tissue. A patient who was planning to undergo surgical debridement of an ulcer in the sacral region that was caused by lying down for a long period after a cerebral hemorrhage was recruited in Fujisawa City Hospital, Kanagawa, Japan. Wound blotting was performed just prior to surgical debridement and the debrided tissue embedded in paraffin. The ulcer, which had a 2.9 cm major axis, was divided into 20 areas approximately 0.35 cm² each, and the consistency of tumor necrosis factor-α positivity between the wound blotting samples and tissue sections was examined in each area. The sensitivity and specificity of wound blotting were 89% and 82%, respectively. This wound blotting method noninvasively revealed the protein distributions within the wound tissue.

Pressure ulcers are a physical and mental burden to patients. In clinical settings, the treatment strategy for ulcers is usually determined by subjective assessment, which can be difficult even for expert medical staff.¹ Therefore, the establishment of an objective assessment tool is required.

  Several studies have recently reported the use of a biochemical analysis of wound exudate for wound assessment.^2-4 Wound exudate can be collected noninvasively and its protein component reflects the wound healing process.⁵ In most of these studies, the pooled exudate was collected by sealing the wound surface with a polyurethane film dressing for several hours and then used for biochemical analysis.^2,5 However, during the time in which the exudate is pooled, the sealing on the wound prevents the use of a dressing which is needed to maintain the appropriate moisture balance. In addition, Minematsu et al⁶ also reported that the protein components of the exudate degraded during the pooling period. These authors recently proposed the wound blotting method used to collect at least a few microliters of fresh wound exudate by attaching a nitrocellulose membrane to the wound surface.⁶ This analysis revealed a significant correlation of the tumor necrosis factor-α (TNF-α) concentration between immunostaining of the blotting membrane and western blotting analysis of the pooled exudate.

  The unique property of the wound blotting method is that it can be used to analyze the distribution of protein components on the wound bed. Previous studies reported the association of TNF-α distribution patterns with wound healing progress.⁶ Tumor necrosis factor-α distribution analysis could be a method for predicting wound healing. In the case of large ulcers, the physiological wound status is often heterogeneous. In fact, some wound status assessment tools, such as DESIGN-R and the Pressure Sore Status Tool, evaluate the distribution of granulation tissue and necrotic tissue on the wound surface.^7,8 However, the consistency of the distribution analysis of wound blotting with actual distribution within wound tissue has not yet been revealed, although the wound blotting method is the only approach for analyzing protein distribution on the wound surface. This study aimed to investigate the protein distribution detection consistency between wound blotting and immunohistochemistry.

A 60-year-old woman diagnosed with cerebral hemorrhage was hospitalized after being found in bed with a decreased consciousness level by her brother. She was taken to Fujisawa City Hospital, Kanagawa, Japan, by ambulance, where a pressure ulcer in the sacral region was found. She received conservative treatment for cerebral hemorrhage. The patient was diagnosed with hemiplegia and spent most of her time in the decubitus position. She also had comorbidities of hypertension and depression. Sharp debridement was performed on the pressure ulcer and cadexomer iodine ointment was used every day. The pressure ulcer area covered with necrotic tissue decreased in size, however, the center of the ulcer remained covered with necrotic tissue as well as granulation tissue that appeared edematous (Figure 1A). The patient underwent surgical debridement and negative pressure wound therapy. After 1 week, granulation tissue formation occurred and the treatment was changed back to ointment.

  The patient and her family read the study protocol and provided written informed consent. The study protocol was approved by the Research Ethics Committee of the Graduate School of Medicine of the University of Tokyo (Tokyo, Japan) and Fujisawa City Hospital, Kanagawa, Japan.

  Wound blotting procedure. Before surgical debridement, fresh exudate was collected using the wound blotting method as previously described,⁶ with slight modifications. Briefly, the ulcer and peripheral skin were cleaned with soap and water, excess water and exudate were absorbed with gauze, and the fresh exudate was collected by attaching nitrocellulose membranes to the wound surface for 10 seconds. Since the wound surface was largely uneven, 16 mm-wide strips of nitrocellulose membranes (Supported Nitrocellulose Membrane, Bio-Rad, Hercules, CA) were used to cover the entire wound surface (Figure 1B). The collected membranes were stored at 4°C until use.

  Following total protein staining (Reversible Protein Stain Kit, Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s instructions, the blotting membrane was incubated with 0.3% hydrogen peroxide in 20% methanol for 30 minutes, and with a blocking agent (Blocking One-P, Nacalai Tesque, Kyoto, Japan) supplemented with 500 µg/mL of levamisole, (Sigma-Aldrich, St. Louis, MO) for 30 minutes. Tumor necrosis factor-α was detected indirectly using goat polyclonal antibody for TNF-α (dilution 1:250; Santa Cruz Biotechnology, Santa Cruz, CA) and alkaline phosphatase-conjugated anti-goat Immunoglobulin G antibody (dilution 1:1000, Anti-Goat IgG Antobidy Agarose Conjugated, Rockland Immunochemicals, Gilbertsville, PA). Immunoreactivities were visualized using a chemiluminescent substrate (DuoLux Chemiluminescent/Fluorescent Substrate for Alkaline Phosphatase, Vector Laboratories, Burlingame, CA), and captured using a chemiluminescence digital imaging apparatus (LumiCube, Liponics, Tokyo, Japan) (Figure 1C).

  Analyzing software (Image J, National Institutes of Health, Bethesda, MD) was used to process the membrane images. All images were initially flipped horizontally. The wound edge was identified from the total protein staining image, and the TNF-α distribution signal within the wound site was extracted. The signals were separated into 3 channels (red, green, and blue), and the image of the blue channel was selected for further processing, including the reduction of noise (< 5 pixels) by median filtering and binarizing at the threshold, which is the mean intensity + 2SD of a 10-point measurement in the skin surrounding the wound (Figure 1D).

  Tissue collection and histological analysis. The debrided tissue, including the entire wound and the surrounding skin (Figure 2A), was fixed in 10% neutral-buffered formalin. The wound tissue was transversely cut into 5 mm wide pieces (labeled 1 through 5) and divided into right and left sides along the median line of the wound (Figure 2B). The tissue pieces were embedded in paraffin and cut into 5 µm thick sections from the caudal plane of each block. The sections were used for hematoxylin and eosin staining and TNF-α immunostaining using the horseradish peroxidase-based avidin-biotin complex method (VECTASTAIN ABC Kit, Vector Laboratories, Burlingame, CA) and 3,3’-diaminobenzidine tetrahydrochloride. Staining was evaluated at the central and peripheral zone of the wound site under stereoscopic (LZ, Kenis, Osaka, Japan) and inverted (BX41, Olympus, Tokyo, Japan) microscopy.

  Analysis of TNF-α distribution. The distributions of TNF-α within the tissue were evaluated (positive or negative) in all 20 areas, as well as within the central and peripheral wound zones of 10 blocks comprised of the left and right sides of pieces 1-5. Similarly, the immunoreactivity for TNF-α was evaluated in the corresponding 20 areas of the blotting membrane. The agreement (sensitivity and specificity) between the immunohistochemistry findings and the immunostaining findings of the blotting membrane was then calculated.

  Macroscopic and histological observations of the ulcer. The pressure ulcer in this case was round, measuring 2.9 cm in its major axis and 2.6 cm in its minor axis. Edematous granulation tissue formed the peripheral area, and soft and pale necrotic tissue was observed at the center of the ulcer (Figure 1A).

  Hematoxylin and eosin staining revealed the infiltration of inflammatory cells, including large numbers of neutrophils and macrophages as well as a few eosinophil granulocytes in most areas. Denaturated collagen was seen on the surface of the central parts of the ulcer. The surrounding area was composed primarily of granulation tissue in which bleeding, nodes of edema, and degenerated fibroblasts were frequently observed (Figures 2C and 2D).

  Immunohistochemistry findings revealed that the inflammatory cells, fibroblasts, vascular endothelial cells, and basal and prickle cell layers of the regenerating epidermis were positive for TNF-α in 9 of the 20 areas of wound tissue (Table 1). Tumor necrosis factor-α-positive cells were the most abundantly observed on the right sides of sample numbers 2 and 3. The TNF-α signals were detected in a relatively deeper layer of granulation tissue on the left side of sample number 1. Most of the immunoreaction for TNF-α was localized in the vascular endothelial cells of the peripheral region of the right side of sample number 1. A few TNF-α-positive cells condensed in a localized area of the wound surface (Figure 2E).

  Protein distribution detected by wound blotting. Total protein staining revealed the composition of wound exudate (Figure 1C). Immunostaining indicated the presence of TNF-α in the upper and right sides of the ulcer. A histogram of the signals within the wound site showed a sharp peak at 12 arbitrary units (au) of intensity similar to that of the surrounding skin as well as a blunt peak around 70 au. The threshold for binarization, which was a mean + 2SD of the surrounding skin area, was appropriately divided into positive and negative for TNF-α. The TNF-α-positive areas consisted of 10 of the 20 areas (50%) of blotting membrane (Figure 1D).

  Concordance rate of immunohistochemistry and wound blotting. A comparison of results between tissue sections and blotting membrane is shown in Table 1. The sensitivity and specificity of the wound blotting method for TNF-α were 89% and 82%, respectively.

Comparison of the 20 areas of the pressure ulcer showed the high consistency of positivity for TNF-α between the wound blotting analysis and immunohistochemical analysis of the tissue, indicating that wound blotting is a promising method to reveal protein distribution within tissue. Wound blotting could be substituted for histological analysis using biopsy samples. Since wound blotting is a noninvasive and repeatable method, it is a suitable assessment tool for pressure ulcers in clinical settings. Therefore, it could aid medical professionals in the timely selection of the most appropriate treatment based on wound tissue status. It is also useful for the continuous evaluation of long-term wound pathophysiology in clinical and research settings.

  Tumor necrosis factor-α released from macrophages, neutrophils, fibroblasts, and keratinocytes is a multifunctional proinflammatory cytokine. The secretion level of TNF-α in chronic wounds is higher than that in acute wounds due to excessive activity of TNF-α converting enzyme (TACE) and/or a deficiency of its inhibitor.⁹ Anti-TNF-α agents accelerate the healing in impaired cutaneous wounds.^10,11 Because TNF-α signals detected by the wound blotting presumably reflect the status of inflammation, TNF-α distribution analysis by wound blotting could aid health care professionals in selecting treatment options for ulcers.

  False-negative TNF-α results were obtained in the right peripheral region of sample number 1. Immunoreactivity for TNF-α was mostly observed in the vascular endothelial cells and epidermal basal and prickle cells in this area. In contrast, numerous inflammatory cells and fibroblasts were positive for TNF-α in addition to endothelial cells in the other regions, in which consistent distributions were detected by both wound blotting and immunohistochemistry. These differences in distributions were probably because of the expression/activity of TACE, which cleaves the transmembrane from of TNF-α and induces its release from cells.¹² Further examination of TACE expression/activity is required to validate the TNF-α results.

  The depth of the localization of positive cells from the wound surface is also one of the factors influencing the detection of TNF-α by wound blotting. In the right area of sample number 2 and 3, in which the strongest signals were detected by wound blotting, many TNF-α-positive cells were observed on the tissue surface. In the left area of sample number 1, where the wound blotting signals were weak, TNF-α tested as strongly positive as that in the right area of sample number 2 and 3, but the TNF-α-positive cells were located in the deeper layer of the tissue. On the other hand, TNF-α signals were slightly detected by wound blotting in the left peripheral region of sample number 1, in which few positive cells were concentrated on the surface layer of the wound bed in a localized area. Therefore, immunoreactivity intensity is an informative value that reflects the depth and number of positive cells, although signal intensity was not evaluated in this study because of binarization. An improved wound blotting analysis method is required.

  Two possible reasons for the false positive result detected in 2 areas were considered: a nonspecific antibody reaction and exudate diffusion. The nonspecific antibody reaction is a common problem of membrane immunostaining in both western blotting and wound blotting. Since the false positive was detected only at the membrane’s edge, a nonspecific antibody reaction cannot be excluded. Total protein staining, however, indicated a greater amount of exudate in these areas. The diffusion of protein on the blotting membrane was revealed in the case of abundant exudate.6 Therefore, the authors believe the false positive result was probably because of the diffusion of TNF-α onto the blotting membrane rather than the nonspecific antibody reaction. A detailed examination of the appropriate attachment time of the membrane to the wound surface according to the amount of exudate may improve the wound blotting method.

  The limitation of this study was that the authors examined the protein distribution in only 1 pressure ulcer. Debrided tissue could be obtained from only 1 ulcer because the inclusion criterion of this study was limited to ulcers debrided as a single piece. The ratio of contracted length to original length to define corresponding areas was used because the debrided tissue contracted; however, the original length of the ulcers debrided in pieces was not known. Therefore, ulcers debrided in pieces were excluded. Further studies are required to compare protein distributions in a large number of pressure ulcers including those with different depths.

This study showed consistency in TNF-α distributions determined by the wound blotting method and immunohistochemistry. Therefore, the wound blotting method could be used as a noninvasive assessment tool for the biochemical analysis of pressure ulcers, including portions with differing physiological statuses.

The authors would like to thank the patient for her cooperation.

Affiliations: Aya Kitamura, MHS; Gojiro Nakagmai, PhD; Mikako Yoshida, PhD; Hiroshi Noguchi, PhD; Yoshimi Nishijima, PhD; Takeo Minematsu, PhD; Taketoshi Mori, PhD; and Hiromi Sanada, PhD, RN, WOCN are from the University of Tokyo, Tokyo, Japan. Ayumi Naito, MHS; Jun Sugawara, MD; Kazuo Takahashi, MD; Junichi Umemoto, MD; Nao Terada, RD; and Ryo Segawa, BP are from Fujisawa City Hospital, Kanagawa, Japan. Hiroki Shibayama, MD is from Saitama Medical Center, Saitama, Japan. Amiko Hakuta, MD is from Yokohama City University Hospital, Kanagawa, Japan.

Correspondence:
Hiromi Sanada, PhD, RN, WOCN
Department of Gerontological Nursing/Wound Care Management
Graduate School of Medicine
University of Tokyo
7-3-1 Hongo, Bunkyo-ku
Tokyo 113-0033, Japan
hsanada-tky@umin.ac.jp

Disclosure: This study was supported in part by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science.