Zijun Zhang, Orthobiologic Laboratory, Union Memorial Hospital, 201 E. University Parkway, Bauernschemidt 778, Baltimore, MD 21218, USA. T: + 1 410 5542830; F: + 1 410 5542289;E: email@example.com
Plantar fat pad (PFP) is a tissue structure that absorbs the initial impact of walking and running and ultimately bears body weight at standing. This study was designed to quantify the histomorphological changes of the PFP in aged rats. The most medial PFP was dissected from the hind feet of young rats (4 months old, n = 6) and aged rats (24 months old, n = 6). Histological structure and cellular senescence of PFP were analyzed stereologically and histomorphometrically. Immunohistochemistry of matrix metalloproteinase 9 (MMP9) was also performed on PFP tissue sections. Compared with young rats, the thickness of epidermis, dermis and septa of the PFP were significantly reduced in the aged rats. The total volume of adipose tissue in the PFP of aged rats was only about 65% of that in the young rats. The microvascular density and the number of fat pad units (FPU), a cluster of adipocytes enclosed by elastin septa, in the PFP were unchanged in the aged rats. In the aged rats, the number of adipocytes per FPU was reduced but the number of simple adipocyte clusters, without surrounding septa, was increased. The shift of the types of adipocyte clusters in the aged PFP was accompanied by degradation of elastin fibers and increased expression of MMP9. In conclusion, the PFP, particularly the elastic septa, degenerates significantly in aged rats and this may contribute to the pathology of PFP-related diseases.
Plantar fat pad (PFP) is a composite tissue structure. From superficial to deep, the PFP consists of the epidermis, dermis, panniculus carnosus muscle, and subcutaneous adipose tissue. Functionally, the PFP bears body weight and cushions the impact on the foot. Computational modeling indicates that the PFP has a non-linear, viscoelastic behavior at variable strain rates (Miller-Young et al. 2002; Fontanella et al. 2012). There are numerous biomechanical models that quantitatively predict the properties of PFP (Erdemir et al. 2009; Chokhandre et al. 2012). In histology, a hallmark of the PFP is clusters of adipocytes enclosed by dense fibrous septa (Jahss et al. 1992), hereafter referred to as fat pad units (FPU). In the PFP, it is believed that the adipocyte clusters absorb compressive stresses that are incurred during daily activities, while the fibrous septa prevent displacement of the adipocytes (Kimani, 1984). The PFP alone is capable of absorbing about 30–60% of deforming energy and that significantly reduces the impact of mechanical loading on other joints such as ankles and knees (Bennett & Ker, 1990; Rchallis et al. 2008).
The physiological importance of the PFP has led to broad speculation of its roles in foot functions and diseases. Heel pain is common and its pathology is predominantly mechanical (Thomas et al. 2010). Body weight, compressibility index and the thickness of the PFP are recognized as factors that are relevant to heel pain (Prichasuk, 1994; Yi et al. 2011), although how precisely those factors contribute to the foot pathology is still a subject of debate (Spears & Miller-Young, 2006). Persistent heel pain appears in patients older than 40 and is associated with reduced elasticity of heel pad (Ozdemir et al. 2004). In certain cases, the thinning or atrophy of PFP is believed to cause heel pain (Thomas et al. 2010; Yi et al. 2011). PFP atrophy and associated pathology occur ubiquitously due to aging and can be accelerated in disease states such as diabetes, rheumatoid arthritis, peripheral vascular disease, and peripheral neuropathy (Buschmann et al. 1995). In a group of patients with unilateral plantar heel pain, the displacement of the symptomatic PFP was significantly less than that of the painless PFP, particularly in the loading range from 0 to 1 kg (Tsai et al. 1999). Although it seems clear that altered mechanical properties of PFP is important for the development of PFP-related foot diseases, the underlying cellular and molecular pathology is still elusive.
Pressure ulcers are serious medical complications in aged populations. The heel is the second most common site for pressure ulcer (Cichowitz et al. 2009). Measured by compression and shear, the PFP in diabetic patients is stiffer than in the healthy controls. In diabetic patients, the PFP at the calcaneus showed decreased energy dissipation. The changed mechanical properties of PFP in diabetic patients are risk factors of foot ulceration (Brink, 1995; Pai & Ledoux, 2010, 2012).
Given the prevalence of PFP-related pathological conditions, the biology and pathology of PFP have not been fully appreciated. A few clinical observations on PFP conflict with each other regarding the role of PFP in heel pain (Prichasuk, 1994; Waldecker & Lehr, 2009; Thomas et al. 2010). The biochemistry and histomorphology of PFP in aging and diabetes have been studied using human cadaveric tissues (Jahss et al. 1992; Waldecker & Lehr, 2009; Wang et al. 2011), which are most relevant to PFP-related diseases but provide only sectional information of the pathology. Furthermore, these studies are limited by the inherent variations of human PFP tissues influenced by daily activities, body weight, genetics, genders, and other co-morbidities (Rchallis et al. 2008). A validated animal model of the PFP, with controlled genetic background, age, and gender, would allow for more vigorous evaluations of the PFP biology and its role in disease states.
This study was designed to characterize PFP in rats and its histomorphologic changes during aging, which has been linked to common heel diseases. The histology of PFP was compared between young rats and aged rats and quantified with stereology. In addition, cellular senescence and the expression of matrix metalloproteinase 9 (MMP9) in the PFP were investigated.
Materials and methods
Six 4-month-old and six 24-month-old male Long–Evans rats were euthanized (approved by Institutional Animal Care and Usage Committee) and the most medial PFP was harvested using sharp dissection from both hind paws. The collected PFPs (12 from young rats and 12 from aged rats) were fixed in 4% paraformaldehyde, washed in phosphate-buffered saline (PBS) and transferred into 25% sucrose solution. The tissue samples were embedded in Tissue-Tek®O.C.T. compound (Sakura, Japan). The orientations of individual specimens were determined using the vertical uniform random (VUR) sectioning technique (Baddeley et al. 1986). PFPs were placed skin-side down in Cryomold® specimen molds (15 × 15 × 5 mm, Sakura), defining the vertical axis as perpendicular to the skin layer. A random number generator was used to provide the initial orientation (0° ≤ x ≤360°) of the first specimen around the vertical axis. Increments of 15° were added to the initial value to determine subsequent random orientations for the remaining 11 specimens from both young and aged rats. A printed 360° compass was placed underneath the clear specimen molds, which contained PFPs and O.C.T. embedding media. The specimens were rotated along its vertical axis to be in line with the angle on the compass that was based on initial randomization and (15°) increment. All specimens were frozen at −80 °C overnight. The PFPs were exhaustively sectioned at a thickness of 10 μm, yielding approximately 300 sections per specimen. Sequential sections were placed on silane-coated slides. An initial slide was randomly chosen using a random number generator and subsequent slides were selected at 20-slide intervals. A set of 15 slides per PFP were used for each histochemical staining.
For observation of the histological structure, hematoxylin and eosin (H&E) staining was performed on tissue sections of PFPs harvested from both young and aged rats. Elastin staining was also conducted by creating a working solution composed of alcoholic hematoxylin, ferric chloride and Weigert's iodide solution, and subsequently differentiating the specimens in ferric chloride until the desired colors were achieved. Collagen was stained with Van Gieson's solution for distinction.
Senescent cells on PFP sections of young and aged rats were detected by the activity of senescence-associated β-galactosidase at pH 6 (Dimri et al. 1995). Sections were rehydrated in PBS for 10 min and incubated with a mixed solution that contained 5-bromo-4-chloro-indolyl-β-d-galactopyranoside (X-gal, Senescence Cells Histochemical Staining Kit; Sigma-Aldrich, St. Louis, MO, USA) overnight at 37 °C without CO2 supplementation.
All the PFP slides that were stained with H&E and for the activity of senescence-associated β-galactosidase were digitized with a whole slide scanner (Aperio, Vista, CA, USA) at 1×, 4×, 5×, 10× and 20×. The images were analyzed stereologically using ImageJ (NIH, Bethesda, MD, USA). The thickness of epidermis and dermis were measured using the orthogonal intercept method (Ferrando et al. 2003; Da Costa et al. 2007). Briefly, parallel lines 100 μm apart were randomly overlaid on each section. Every intersection with the epidermis, dermis, and elastic septa was counted and an orthogonal intercept was drawn to the structure of interest (Fig. 1). The arithmetic thickness was calculated from the measurements.
The Cavalieri principle was used to estimate the adipose tissue volume in PFP (Mouton, 2002). A 100 × 100 μm grid was randomly overlaid on each section. The boundaries of the fat tissue in PFP were identified with a mask and each intersection of the probe within the determined fat tissue boundary was counted. Adipose tissue volume was estimated using the following formula:
where V is the volume estimate, d is the distance between sections at 200 μm, t is the section thickness at 10 μm, a(p) is the area corresponding to each point determined by the grid density 1000 μm2, and ΣP is the sum of points counted within the boundary of interest.
Using the physical dissector probe, cellular senescence was quantified by randomly overlaying a 50 × 50 μm counting probe in a systematic uniform manner using the following formula.
where Nv is the number of senescent cells per unit volume, ΣQ is the sum of the objects counted, n is the number of dissectors sampled, a(p) is the area of the dissector probe used at 2500 μm2, and h is the height of dissector probe at 10 μm.
Collections of adipocytes and microvasculature contained within elastic septa were observed as fat pad units (FPU). These FPUs were quantified by absolute manual counting due to their low frequency per slide.
Two slides stained with H&E were randomly selected from each specimen. The numbers of adipocytes per FPU, which was designated as collections of adipocytes surrounded by fibrous septa, as well as the number of adipocyte clusters that consisted of more than three adipocytes and did not have a fibrous tissue wall were counted using a threshold mask in Adobe Photoshop (Adobe, San Jose, CA, USA). Simultaneously, the longest dimension of adipocytes (n = 177 for young rats and 160 for aged rats) was measured in the PFP of both young and aged rats.
Immunohistochemistry of matrix metalloproteinase 9
From each PFP, two slides were randomly selected and rehydrated in 100 mm Tris-buffered saline (TBS) with 0.025% Triton X-100 for 10 min, followed by blocking with 10% normal rabbit serum in 100 mm TBS with 1% bovine serum albumin. Incubation was carried out with goat anti-rat MMP9 antibody (C-20, 1 : 100 dilution; Santa Cruz Biotechnology, Dallas, TX, USA) overnight at 4 °C and secondary biotinylated rabbit anti-goat IgG for 30 min (Vector Laboratories, Burlingame, CA, USA). After incubated with avidin conjugated horseradish peroxidase for 30 min, slides were developed in 3,3′-diaminobenzidine. The staining of MMP9 in PFPs was viewed under a light microscope.
Data is presented as mean ± SD. Student's t-test was used to compare the stereological data of PFP between young and aged rats. P <0.05 was considered statistically significant.
On histology, the superficial layers, epidermis and dermis, of PFP showed no structural difference between young and aged rats. In the subcutaneous layer, clusters of adipocytes were enclosed by dense fibrous septa, forming the typical FPUs. In FPU, there were microvascular bundles around adipocytes. There were also adipocytes that clustered together without a septal wall (Fig. 2).
Stereological evaluation revealed significant changes in the aged rats in the thickness of epidermis, dermis, and elastic septa of PFP. The thickness of epidermis of PFP in aged rats was decreased by approximately 10% compared with that in young rats (390.0 ± 96.1 μm in young rats vs. 351.0 ± 115.0 μm in aged rats; P = 0.032). Similarly, the thickness of dermis of PFP in the aged rats was decreased by 30.8% as compared with young rats (168.0 ± 4.2 vs. 242.8 ± 94.2 μm, P = 0.015). The thickness of septa in the PFP of aged rats was decreased by 24.4% (19.0 ± 9.0 μm in young rats vs. 14.3 ± 4.9 μm in aged rats; P = 0.024; Fig. 3).
When the total volume of adipose tissue in the PFP was compared between the young and aged rats, it was significantly less in aged rats (3111 ± 118 μm3) than in young rats (4719 ± 121.3 μm3; P = 0.048; Fig. 4A). However, there were no differences in the number of FPUs and microvascular density between the PFPs from young and aged rats (Fig. 4B,C). The mean number of adipocytes per FPU was significantly less in the aged rats compared with the young (7.3 ± 3.8 vs. 14.5 ± 7.6; P < 0.001; Fig. 5A). Except for intact FPU, there were also adipocyte clusters defined as containing more than three adipocytes and not associated with a neurovascular bundle in PFP. The number of adipocyte clusters in the PFP was significantly greater in the aged rats (18.1 ± 2.35) than in the young rats (13.4 ± 2; P < 0.05; Fig. 5B). The longest dimension of the adipocytes in the PFP was 48.3 ± 32.5 μm for the young rats and 41.2 ± 25.1 μm for the aged rats (P > 0.05; Fig. 5C).
Senescence-associated β-galactosidase was detected in the PFP of both young and aged rats. Furthermore, most of the staining was in fibroblasts distributed throughout the dermis (Fig. 6A). Although there was a trend for aged rats to undergo greater cellular senescence (1115 ± 121 cells mm−3 in aged rats vs. 997 ± 116 cells mm−3 in young rats), this was not statistically significant (P = 0.21, Fig. 6B). Other tissue structures underneath PFP, including muscle, tendon, and bone, did not show staining for cellular senescence (data not shown).
The morphology of elastic fibers showed remarkable differences. Whereas elastic fibers branched out and formed networks in the young rats, they appeared dense and lost fine branches in the aged rats (Fig. 7). The PFP in aged rats demonstrated intense MMP9 staining, whereas MMP9 was minimally stained in the PFP in young rats. MMP9 in the PFP of aged rats was mainly localized around capillary structures in the FPUs and at the junction between epidermis and dermis (Fig. 8).
Plantar fat pad properties change as aging progresses (Kinoshita et al. 1996; Ozdemir et al. 2004). The biology of PFP aging, however, is poorly understood. While PFP shares common aging processes with other tissues in the body, its weight-bearing nature adds another layer of complexity to the aging of PFP. To establish an animal model for studying PFP, this study quantified the basic structural features of PFP in aged rats.
Stereologic analyses found that the thickness of epidermis, dermis and septa in the PFP were significantly reduced in aged rats. This finding is largely consistent with skin thinning commonly developed during aging, as a result of reduced collagens, glycosaminoglycans and water in the extracellular matrix of skin tissue (Calleja-Agius et al. 2007). However, the biology of skin aging is sophisticated and is topically specific (Makrantonaki & Zouboulis, 2007). This study and others (Thomas, 2005) showed a greater reduction in thickness of the dermis compared with the epidermis in the PFP of aged rats, but this was not seen in the aged skin of other parts of the body (Lavker et al. 1989; Thomas, 2005). While the irradiation of ultraviolet light is largely responsible for the aging of skin covering most parts of the body (Makrantonaki & Zouboulis, 2007), accumulative mechanical stress may play a role in the aging of PFP, similar to the degeneration of articular cartilage (Beckett et al. 2012). Conversely, aging of the PFP inevitably changes its mechanical properties and functionality.
Typical FPUs, which are a cluster of adipocytes enclosed by dense fibrous septa, presented in the PFP of both young and aged rats. In general, fat tissue redistributes and fat mass decreases during aging (Tchkonia et al. 2010). However, fat tissue is also regulated in a region-dependent manner (Cousin et al. 1993; Tchoukalova et al. 2010). In the aged rats, the total volume of adipose tissue in PFP was reduced as well as the thickness of the elastic septa, which confine adipocytes in clusters. It has been found that there are strong positive correlations between age and stiffness of the plantar soft tissues in several areas of the feet (Kwan et al. 2010). The amount of compressibility that is lost due to the reduced volume of adipose tissue and the reduced thickness of the elastic septa during aging remains to be quantified. It is certain, however, that progression of aging in the PFP will alter the mechanical properties of the PFP and as a result change biomechanical loading of other neighboring joints in the lower extremity. The PFP is often involved in complications of the diabetic foot. Interestingly, it appears that the thickness of the elastic septa and the dermis are increased in diabetic PFP (Wang et al. 2011). Thinning of the epidermis and stiffening of plantar soft tissues in diabetic PFP increase the risk of tissue breakdown and ulcer formation (Chao et al. 2011). Although the biology of aging and the pathology of diabetes are often intertwined, each has featured structures and properties of PFP.
The FPU is essential for the weight-bearing and shock-absorbing functions of the PFP (Jahss et al. 1992). Using stereology, this study confirmed that the number of FPUs in the PFP was not changed in aged rats. The number of adipocytes per FPU in the aged rats, however, was decreased by approximately 50%, as compared with the young rats. This difference may well characterize FPU in aging, but it does not represent a reduction of the total number of adipocytes in the PFP. Studies on the diabetic PFP have found that the total number of adipocytes remains unchanged but the volume of adipocytes may or may not be reduced (Waldecker, 2001; Wang et al. 2011). The FPU was featured as a collection of adipocyte clusters and a neurovascular bundle encased by an elastic septum. In the PFP, there were also simply adipocyte clusters (> three adipocytes) without well-defined septa. The number of adipocyte clusters was increased in the aged rats. This finding suggests that the aging process causes redistribution of the forms of chondrocyte clusters in PFP: it reduced the number of adipocytes per FPU but increased the number of adipocytes in the form of simple clusters. It is reasonable to speculate that the simple adipocyte clusters would not have comparable mechanical properties as FPUs. The diameters of adipocytes varied greatly in the PFPs of both young and aged rats. Therefore, the size of adipocytes is unlikely a factor of PFP aging.
In an FPU, the septum is composed primarily of elastin (Buschmann et al. 1995), which gives the PFP added viscoelastic qualities and aids its function as a shock absorber. Elastic septa in diabetic PFP were found to degrade into segments (Buschmann et al. 1995). The thickness of septa was reduced in the PFP of aged rats. In addition, noticeably less elastin content surrounding the total adipose tissue as well as in the septal walls was observed in the PFP of aged rats. The elastin fibers of the aged rats were disorganized and piecemeal in appearance, signs of degradation. MMP9 is expressed by endothelial cells and degrades elastin and other collagens (Lau et al. 2008). Adipocytes also express MMP9 and their expression is up-regulated in response to inflammation (O'Hara et al. 2009). Immunohistochemistry of MMP9 showed an increased expression in the FPUs, specifically in the microvascular bundles, and epidermis in the aged PFP. Increased MMP9 activity at the junction of epidermis and dermis may accelerate matrix breakdown and contribute to the reduced thickness of epidermis and dermis in the aged PFP. The increased MMP9 activity in the FPUs of aged rats may be largely responsible for the reduced thickness of septa and elastin degradation. In the aged PFP, MMP9 was mainly localized to the endothelial cells of the capillaries surrounded by adipocytes. This suggests that, other than adipocytes and septa, microvascular bundles and vascularity are also important biological elements of PFP. The stereologic data demonstrated that vascular density in PFP was similar between young and old rats, although it is widely suspected that vascular insufficiency is a probable cause of plantar pad atrophy in systemic disease states such as diabetes and peripheral vascular disease. Degenerative septa could lead to the shift of types of adipocyte clusters in the aged PFP and be detrimental to the functionality of the PFP due to impaired capability to redistribute the impact energy from daily activities.
Cellular senescence is an important part of aging biology (Naylor et al. 2013). In aging studies, the activity of senescence-associated β-galactosidase has been widely used to determine aging cells. This method is convenient to perform and can identify senescent cells from a heterogeneous cell population (Shlush et al. 2011). The number of cells with senescence-associated β-galactosidase activity in the PFPs of aged rats was greater than that in young rats, but this did not reach statistical significance. Although stereology offers a means of quantification of senescent cells throughout the PFP, the activity/staining of senescence-associated β-galactosidase may be affected by the variables introduced during the staining procedure. In addition, it has been suggested that a group of biomarkers, rather than a single marker, are required to define the aging-related senescence (Sikora et al. 2011). Nevertheless, stereologic analysis of aged PFP demonstrated overwhelmingly tissue degeneration, such as reduced tissue volume and increased MMP activity.
Fat pads exist in many parts of the body. The infrapatellar fat pad (IPFP) is an excellent candidate for comparison with the PFP, because the IPFP experiences similar cyclical loading during the movement of the knees. The histology of the IFPF, apart from the presence of adipocytes, bears little resemblance to that of the PFP (Okura et al. 2008; Clements et al. 2009). The IPFP is a seemingly random collection of adipocytes distributed in the loose connective tissue. It lacks the specialized organization of adipocytes with accompanying neural and vascular structures contained and separated by septa as in the PFP, observed in this study and by Pai & Ledoux (2012). The histologic organization is responsible for the mechanical properties of the PFP (Jahss et al. 1992) and distinguishes the PFP from other fat pads in the body – even ones that undergo cyclical loading. Due to these stark differences between the PFP and other fat pads in histology, the findings of this study are limited to the aging of PFP. Interestingly, MMP9 expression was also increased in the IPFP in osteoarthritic joints (Gandhi et al. 2011). This corroborates that adipose tissue in PFP and IPFP can be a source of inflammatory mediators in states of pathology.
In conclusion, PFP undergoes significant changes structurally and potentially functionally during aging. The degradation of elastic fibers and other matrix components may be responsible for the reduced thickness of epidermis and dermis, and fat tissue volume, and may cause reorganization of adipocyte clusters in the aged PFP. The PFP used in this study and human heel pad shares common physiological functionality and histological structures. The quantifications of tissue structure and uncovered degenerative changes in the PFP of aged rats provide insights into the aging process of human heel pad and may contribute to the understanding of the pathology of PFP-related diseases.
The authors thank Robert McMahan, Johns Hopkins University, for kindly providing rat tissue samples and Dr. Bruce Fenderson, Thomas Jefferson University, for assistance in digitalization of tissue slides. The study was sponsored in part by the Dr. Lew Schon Innovation Fund. The authors declare no conflicts of interest.