Increased Substance P Immunoreactivity in Ipsilateral Knee Cartilage of Rats Exposed to Lumbar Spine Injury

Abstract

Augmented expression of substance P (SP) in the cartilaginous tissue has been shown to promote degenerative changes in the cartilaginous matrix of distal contralateral articular joints via neurogenic inflammation post monoarthritis-induction contributing to the symmetrical spread of osteoarthritis (OA). However, no studies have explored whether similar changes are also present within neurosegmentally linked ipsilateral heterologous cartilage.

Objective

The present study aimed to investigate whether experimentally induced lumbar facet-joint OA lead to degenerative changes and enhanced SP expression within the ipsilateral neurosegmentally linked tibiofemoral cartilage.

Methods

Adult male Sprague-Dawley rats were assigned to left side L5-L6 facet mechanical compression injury (surgery) (n = 6), L5-L6 facet exposure with no compression (sham) (n = 5), or naïve (no surgery) (n = 4) groups. The morphology of the tibiofemoral articular cartilage was assessed using a modified Mankin scoring system. Immunohistochemistry was used to examine the density of chondrocytes stained positive for SP (cells/cm²) in the ipsilateral tibiofemoral cartilage at 28 days postintervention.

Results

Tibiofemoral cartilage in the surgery group showed consistent loss of superficial zone chondrocytes, mild roughening of the articular surface and occasional chondrocyte clusters as well as a greater density of SP mainly in the superficial cartilage zone compared with sham and naïve groups, although they also had a basic SP-expression.

Conclusion

Our results support the hypothesis that neurogenic mechanisms may mediate the spread of SP to neurosegmentally linked heterologous joints affecting the distal cartilage homeostasis. These findings contribute additional insight into the potential role of neurogenic inflammation with implications in the pathophysiology of chronic inflammatory joint disease and OA.

Introduction

Osteoarthritis (OA) is a prevalent, painful, and progressive musculoskeletal condition that is characterized by the disruption of the joint microenvironment, affecting the cartilage and surrounding tissues, including subchondral bone, synovium, ligaments, and muscles.^1,2 OA affects synovial articular joints, with the highest prevalence observed in the knee, hand, hip, and spine.³ A number of systemic risk factors for OA have been described, including age, gender, ethnicity, hormones, genetics, obesity, and nutrition, as well as biomechanical factors, including prior injury, muscle weakness, joint deformity, and ligament laxity.^1,4-6 Despite the growing awareness of these risk factors, the pathophysiology of OA is still unresolved. Emerging clinical and experimental research suggests that the nervous system may play an essential role in mediating the pathophysiology and symmetrical manifestation of OA.^4,5,7,8

OA has been traditionally considered an asymmetrical disease; however, studies have shown a predilection over time in humans toward symmetrical involvement, especially in knees⁹ and hands.¹⁰ Research shows that experimentally induced unilateral knee OA in animals results in symmetrical spread of neurogenically mediated inflammation and articular degeneration to contralateral homologous joints.^11-13 Spine facet joints are regularly associated with back pain radiating to lower extremities such as knee and foot¹⁴; however, no animal studies of lumbar facet joint injury exist addressing similar mechanisms of spread of neurogenically mediated inflammation to these joints. Neurogenic inflammation is an acute inflammatory response triggered by the nervous system,¹⁵ which manifests in neurosegmental innervation patterns to release substance P (SP) into peripheral joints and tissues.^15,16 SP is a potent vasodilator and proinflammatory neuropeptide,^17,18 which may influence the pathophysiology of OA through its dose-dependent effect on intraarticular inflammation.^19-21

The specific role of neurogenic inflammation in the pathophysiology of OA is unknown; however, the accumulating research suggests it may be an important mechanism in mediating, and possibly initiating, the systematic spread of OA via neurosegmental patterns.^22,23 The clinical and experimental literature consistently demonstrates that SP is released via peripheral nerve terminals mediating neurogenic inflammation within neurosegmentally innervated homologous joints.^4,13,20,24 In addition, other neuropeptides, via nerve sprouting within cartilaginous tissue (superficial and deep cartilage zone), may be released by peripheral nerves and subsequently worsen OA progression.^25,26 In contrast, SP expression has been found within nonneuronal tissues such as in osteoclasts, macrophages, and chondrocytes.^27,28 Chondrocytes are cartilage cells organized in well-defined tissue zones that maintain the cartilage integrity and cartilage function protecting bones from shearing stress and compressive forces between bone to bone within synovial joints.^2,29 Interestingly, it was previously showed that SP stimulates the production of prostaglandin and collagenases in superficial chondrocytes.²⁹ Prostaglandins play an important homeostatic role in chondrocyte remodeling; however, in OA its abnormal increase leads to cartilage degradation.³⁰ Although previous research demonstrated that SP participates in the joint and cartilage inflammation, no studies have investigated whether its expression is changed within ipsilateral neurosegmentally linked heterologous cartilage joint tissue following induction of spine OA.

The primary purpose of our study was to investigate whether experimentally induced spine OA modulates proinflammatory SP responses within neurosegmentally linked heterologous joint cartilage tissue and if it leads to cartilaginous OA-like changes. In addition, this study aimed to investigate whether a difference in the SP expression between cartilage zone is presented. We set out to test the hypothesis that the concentration of SP expressed within the knee articular cartilage of rats exposed to surgically induced L5-L6 spinal facet joint OA will be greater than rats exposed to either sham intervention (L5-L6) or naive controls changing the cartilage homeostasis. We also hypothesized that the concentration of SP expressed within the superficial zone of rats exposed to surgically induced L5-L6 spinal facet joint OA will be larger than sham intervention (L5-L6) or naive controls.

The findings of this study will address a significant gap in our understanding of the underlying mechanism of SP spread by exploring whether neurogenic mechanisms facilitate the expression of proinflammatory neuropeptides within neurosegmentally linked heterologous tissues. These findings will inform future research investigating mechanisms potentially contributing to the pathophysiology and clinical manifestation of OA.

Methods

Experimental Groups

All animal procedures in this study were approved by the Animal Care Committee of the University of Guelph. Fifteen male Sprague-Dawley rats (303.2 ± 38.3 g) were housed in a room with a stable temperature (23.0°C ± 1.0°C) and fed a regular pellet diet ad libitum. Animals were then randomized into 1 of 3 groups, including surgery (n = 6), sham (n = 5), and control (n = 4).

Induction of Facet Injury

A single injection of carprofen (5 mg/kg body mass) was administered subcutaneously 30 minutes before surgery. Surgery and sham groups were anesthetized using isofluorane and injected subcutaneously over the incision site with a 50/50 lidocaine/maracaine mixture (2 mg/kg body mass). Once animals were anesthetized, a posterior midline incision was made from the spinous processes of L3 to S1 through the skin and subcutaneous tissue. An incision was then made through the thoracolumbar fascia and the erector spinae aponeurosis lateral to the left multifidus muscle. Blunt dissection was used to separate the erector spinae and multifidus muscles. The multifidus tendon attaching to the L5-L6 facet was cut to expose the facet joint. In surgery animals only, the left L5-L6 facet joint was then injured by compressing the L5/L6 facet joint (~1 mm gap in locked position) for three minutes with modified Kelly forceps. This procedure has previously been shown to evoke spine facet cartilage degeneration, tactile allodynia, and spinal cord sensitization 28 days postsurgery.¹⁹ All muscles of surgery and sham animals were then sutured (braided 4-0 coated Vicryl) and the skin was closed using stainless steel skin staples. After regaining consciousness, the rats were returned to their cages and were maintained in the same conditions described above. Naïve control animals did not undergo any surgical interventions and were freely maintained in their cages under the same conditions as surgery and sham animals.

Tissue Preparation, Histology, and Immunohistochemistry

Rats were euthanized by carbon dioxide and their L5-L6 lumbar vertebral segments and left knees were harvested 28 days postsurgery. Lumbar vertebral segments were fixed (10% formalin) for 48 hours and decalcified (Cal-X II, Fisher Scientific, Nepean, Ontario, Canada) for 5 days. L5-L6 was then cut in the transverse plane using a surgical stainless-steel blade and then dehydrated and embedded in paraffin wax for sectioning in the coronal plane. Five-micrometer thick sections were sliced and mounted on microscope slides (Superfrost Gold Adhesion slides, Fisher Scientific, Nepean, Ontario, Canada). Safranin-O staining was performed for visualization of the L5-L6 facet morphology and cartilage degeneration.

Knee tissues collected from the rats were fixed (10% formalin) for 48 hours, decalcified (Cal-X II, Fisher Scientific, Nepean, Ontario, Canada) for 5 days and cut in half in the sagittal plane by a surgical stainless-steel blade. These halves were then dehydrated and embedded in paraffin-wax for sectioning in the coronal plane. Only the medial half was considered for additional analysis since the medial aspect of the tibiofemoral joint has been more associated to cartilage loss and OA development.³¹ Five-micrometer thick sections were cut and mounted on microscope slides (Superfrost Gold Adhesion slides, Fisher Scientific, Nepean, Ontario, Canada) and left unstained for immunostaining or stained with safranin-O/fast green and hematoxylin and eosin (H&E). Tissue immunohistochemistry was conducted using the avidin biotin complex (ABC) method to determine the expression of SP in knee cartilage. The slides were exposed to a series of xylene and alcohol changes to deparaffinize and rehydrate the sections. Sections then underwent antigen retrieval (proteinase K, cat#P6556- 20 µg/ml-Sigma-Aldrich, St. Louis, MO, USA) to unmask antigens and epitopes in order to enhance the staining intensity of the antibody. Blocking solutions (peroxidase blocking solution and DAKO protein block) were applied prior to primary antibody incubation for 45 minutes (1:100 dilution, polyclonal rabbit Anti-Substance P, cat# 20064, Immunostar, Hudson, WI, USA,). A secondary antibody was then applied for 30 minutes (EnVision+TM, Peroxidase, Anti-Rabbit, DAKO Laboratories, Carpinteria, CA, USA) and a 3,3′-diaminobenzidine peroxidase substrate (DAKO Laboratories, Carpinteria, CA, USA) was used for color detection. All steps were applied to negative control slides except the incubation of the primary antibody which was replaced with PBS. Slides were visualized by light microscopy (200× magnification) and imaged with an Olympus BX 60 camera (Olympus, Center Valley, PA, USA). Images were captured and positively stained (brown) cells were manually counted at 200x magnification using Northern Eclipse software (Northern Eclipse v8.0, Empix Imaging, Mississauga, Ontario, Canada).

Data Analysis

To examine the effect of intervention (surgery, sham, control) on the overall SP expression within the tibiofemoral joint cartilage, the density of SP-positive chondrocytes within each of the tibial and femoral cartilage surfaces was calculated ( Fig. 1A-C ) by averaging the density from 3 different regions within each articular surface (tibial, femoral). To determine the density of SP-positive chondrocytes, a grid comprising 100 × 100 µm squares (1 cm²) per grid window ( Fig. 1D ) was superimposed onto the cartilage image (200× field)³² ( Fig. 1C ). The number of SP-positive (brown) chondrocytes was manually counted with the assistance of the Northern Eclipse software within the tibia and femur. The density of SP-positive cells was then calculated based on the number of SP-positive chondrocytes divided by the total cartilage area (µm²) in each cartilaginous tissue. Finally, the total density of the tibial and the femoral cartilages was averaged and recorded as the total tibiofemoral joint cartilage density of SP per animal. To find the total cartilage area, cartilage thickness was multiplied by its length. Cartilage thickness was measured from the articular surface to the subchondral bone and the length measure of 500 µm along the surface was used as a constant measure through all images ( Fig. 1D ). All density values are reported per grid window (cells/cm²). Safranin-O- and H&E-stained knee joint histological sections also underwent histological evaluation for osteoarthritic changes using the modified Mankin scoring system.³³

Figure 1.

Immunostaining images from coronal sections of the rat knee from control, sham, and surgery groups. (A) An overview representation of the knee positively stained with substance P (SP). (B) and (C) represent positively stained SP in the tibial and femoral cartilages respectively, followed by the tibial and femoral negative control images of the surgery group (SP primary antibody was replaced with phosphate buffered saline (PBS) for negative control). (D) A grid 100 × 100 µm (1 cm²) was superimposed onto the image. The number of SP-positive chondrocytes was manually counted through the area of interest (500 µm in length × cartilage thickness) and the positively SP chondrocyte density was then established (number of positive SP cells divided by 500 µm length × full cartilage thickness). In addition, (D) shows the 3 cartilage zones: superficial (S), middle (M), and deep (D) representing 14%, 51%, and 35% of the full cartilage thickness, respectively. A = 100× magnification, B and C = 200× magnification. Scale bars A = 125 µm; B and C = 30 µm.

To inspect for the difference between cartilage zones approximate boundaries of the superficial, middle, and deep zones were identified using a cartilage schematic.³⁴ A grid comprising 100 × 100 µm squares (1 cm²) per grid window was superimposed onto the cartilage image (200× magnification) ( Fig. 1D ). The total cartilage thickness was first determined, and then based on the schematic, proportions occupied by each zone were measured using the ruler tool from the Northern Eclipse software. Thus, superficial zone represented 14% of the overall cartilage thickness; middle zone 51% and deep zone 35%. Positively stained SP chondrocytes within each zone were counted with the assistance of the Northern Eclipse software from tibial and femoral cartilages. The density within each zone was further calculated as the number of SP-positive chondrocytes divided by individual zone area (µm²) that was established as the zone thickness (superficial, middle, and deep) multiplied by the 500 µm length of the surface ( Fig. 1D ). Finally, the density of SP in the tibia and femur in each individual cartilage zone were pulled together (i.e., matching superficial zone of the femur and tibia per animal), averaged and recorded as the tibiofemoral joint density per zone per animal. All SP density values of the superficial, middle and deep cartilage zones were reported as the number of positive cells per cm² (cells/cm²).

Statistical Analysis

To address our main hypothesis, we conducted a 2-way analysis of variance (ANOVA) using the factors of intervention (surgery, sham, control) and location (tibia, femur) with post hoc comparisons using Tukey’s multiple comparison’s test to examine for differences in number of positively stained cells between location and intervention. To address our second hypothesis, we conducted a 2-way ANOVA using the factors intervention (surgery, sham, control) and cartilage zone (superficial, middle, and deep zones) with post hoc comparisons using Sidak’s multiple comparison’s test to examine for the differences in the number of positively stained cells in each individual zone between groups. Tukey’s multiple comparisons was also used to test for differences in number of positively stained cells in the individual zones within each group.

Mankin scores were compared between the three groups using the Kruskal-Wallis variance analysis. Dunn’s multiple comparisons test was performed when P < 0.05. All statistical analyses were performed in Prism (V 7.03). Alpha was set at 0.05.

Results

The L5-L6 facet joints from surgery and sham groups were analyzed histologically to evaluate the impact of the joint compression on facet morphology ( Fig. 2 ). On surgery animals, the left compressed side showed some alteration on the facet cartilage such as loss of proteoglycan staining, some proliferation of chondrocytes in the superficial and upper mid zone and focal retention of matrix staining around chondrocyte clusters in the superficial zone. However, these changes were not presented on all animals. Sham animals demonstrated intact facet joint with healthy morphological features.

Figure 2.

Safranin-O staining on surgery (A, B) and sham (C, D), injured (left) (A, C) and contralateral (right) (B, D) L5-L6 lumbar facet joints. (A) Left injured side showing loss of proteoglycan staining, some articular margin abnormalities, extracellular tissue remodeling and chondrocyte clustering between mid and superficial zone. (B) Contralateral side showing normal cartilage surface with normal articular margins. (C and D) Left and contralateral sham L5-L6 showing smooth and intact cartilage surfaces and normal articular margins (100× magnification, scale bars = 200 µm).

All femoral cartilage samples were analyzed. Two tibial samples in the surgery group as well as one tibia in the sham group were excluded due to structural damage during dissection. Mankin scores of the ipsilateral (left) knees in the spinal control, sham, and surgery groups all showed varying levels of mild safranin-O stain loss and occasional loss of superficial cells. Knee joints in the surgery group had consistent loss of superficial zone chondrocytes, mild roughening of the articular surface and occasional chondrocyte clusters. The modified Mankin scoring of the surgery group (4.4 ± 1.7, n = 6) was significantly different from control (0.6 ± 0.8, n = 4, P = 0.01, Kruskal-Wallis test) and sham (0.6 ± 0.5, n = 5, P < 0.01, Kruskal-Wallis test) groups (mean ± SD) ( Fig. 3 ). No significant difference was found between sham and control (P = 0.99, Kruskal-Wallis test).

Figure 3.

Modified Mankin score of the knee as a function of spinal surgical intervention. Modified Mankin scores were used with 0 indicating normal and a maximum of 13 points indicating end-stage osteoarthritis. Values are presented as means with standard deviation bars. Different letters indicate statistically different groups (P ⩽ 0.01). Significance was set at alpha of 0.05.

The mean and standard deviation of the density of SP-positive chondrocytes/grid window (1 cm²) by location and intervention are reported in Table 1 .

Table 1.

Density of SP Positively Stained Knee Chondrocytes/Grid Window (10,000 µm² = 1 cm²) by Location and Spinal Surgical Intervention.^a

	Control			Sham			Surgery
	Mean	SD	n	Mean	SD	n	Mean	SD	n
Femur	8.74	1.86	4	9.22	0.86	5	11.53	1.96	6
Tibia	8.39	0.86	4	9.10	2.13	4	10.26	0.66	4

^aThe number of substance P (SP)-positive chondrocytes were counted in 3 separate regions within both femoral and tibial cartilage regions in the left knee of rats. Values are expressed as mean and standard deviation.

Additionally, the mean and standard deviation of the density of SP-positive chondrocytes (number of cells/cm²) in the superficial, middle, and deep cartilage zones are listed in Table 2 .

Table 2.

Density of SP Positively Stained Knee Chondrocytes within the Superficial, Middle, and Deep Cartilage Zones as a Function of Spinal Surgical Intervention.^a

	Control			Sham			Surgery
	Mean	SD	n	Mean	SD	n	Mean	SD	n
Superficial zone	12.23	1.05	4	11.03	2.87	5	17.64	2.37	6
Middle zone	7.59	1.00	4	7.66	2.02	5	9.02	1.33	6
Deep zone	8.21	1.18	4	5.99	1.92	5	5.50	1.25	6

^aThe density of the tibial and femoral substance P (SP)-positive chondrocytes within each zone was averaged in the left knee of rats over all groups and was reported as the number of positive cells per cm². Values are expressed as mean and standard deviation.

A 2-way ANOVA revealed a significant effect of spinal intervention on the SP-positive chondrocytes within the knee cartilage of rats, F(2, 21) = 5.563, P = 0.0115. No effect of location, F(1, 21) = 0.9436, P = 0.3424, or interaction between intervention and location was observed, F(2, 21) = 0.3616, P = 0.7008. Post hoc individual comparisons between groups revealed significant increases in SP-positive chondrocytes in the surgery group when compared with both sham surgery (P = 0.0407) and controls (P = 0.0079). No difference was observed between sham and control groups (P = 0.7025) ( Fig. 4 ).

Figure 4.

Density of substance P (SP)-positive knee chondrocytes as a function of spinal surgical intervention. A grid with windows 100 × 100 µm (1 cm²) was superimposed onto the cartilage image (200× magnification). The number of SP-positive chondrocytes was manually counted from three separate regions from each cartilage surface (tibial, femoral). The density of SP-positive cells within each of the tibial and femoral cartilage surfaces was determined by averaging the density of three different regions within each surface. Values are presented as means with standard deviation bars. Significance was set at alpha of 0.05 and denoted by *.

Furthermore, a significant difference in SP-positive chondrocytes between different cartilages zones, F(2, 36) = 67.74, P < 0.0001, as well as an interaction effect between spinal intervention and cartilage zones, F(4, 36) = 9.48, P < 0.0001, was found. The spine facet injury L4-L5 performed in the surgery groups affected the density of the SP-positive chondrocytes within the superficial cartilage zone compared with sham (mean difference [MD] = −6.614; 95% CI [−10.29 to −2.941]; P < 0.0001) and control (MD = −5.409; 95% CI [−8.855 to −1.964]; P = 0.0003) groups. In contrast, no difference within the superficial zone between sham and control (MD = 1.204; 95% CI [−2.613 to 5.021]; P = 0.9789) groups was present. A significant difference within the spinal facet joint-OA surgery group between the superficial zone compared with the middle (MD = 8.619; 95% CI [5.334 to 11.9]; P < 0.0001) and the deep cartilage zones (MD = 12.14; 95% CI [8.854 to 15.42]; P < 0.0001) was evident. In addition, there was difference between the middle and deep zones (MD = 3.52; 95% CI [0.235 to 6.806]; P < 0.0280) within this same group. Similarly, significant difference within the control group between the superficial zone compared to the middle (MD = 4.646; 95% CI [1.047 to 8.245]; P = 0.0040) and the deep cartilage zones (MD = 4.019; 95% CI [4199 to 7.618]; P = 0.0191) was present. A difference between superficial and deep zones (MD = 5.034; 95% CI [1.011 to 9.058]; P = 0.0058) within the sham group was also present but no difference between superficial and middle zone (MD = 3.366; 95% CI [−0.6572 to 7.39]; P = 0.1637) was found ( Fig. 5 ).

Figure 5.

Density of substance P (SP)-positive chondrocytes in knee cartilage zones as a function of spinal surgical intervention. Density was determined as the number of positive chondrocytes within each individual zone divided by the corresponding zone area. Data are represented as mean ± standard error of the mean (SEM). * represents significant difference between groups within the superficial cartilage zone. # represents significant difference between superficial zone compared with corresponding middle and deep zones. “a” denotes a significative difference between the superficial zone to the deep zone within the sham group. “b” denotes a significative difference between middle and deep zones within the surgery group. Significance was set at alpha of 0.05.

Discussion

The results of our study support our hypothesis that the expression of SP is increased within the ipsilateral knee cartilage of rats exposed to experimentally induced left L5-L6 spinal facet joint OA injury, compared with sham and control animals. No difference in SP expression was observed between sham surgery and control animals. Similarly, no difference of SP expression between tibial and femoral cartilage surfaces were found. In addition, our results support the second hypothesis that the expression of SP is increased within the superficial cartilage zone of rats exposed to experimentally induced left L5-L6 facet joint OA injury, compared with sham and control animals. These are the first data demonstrating a causal association between the spread of SP to neurosegmentally linked heterologous tissues following experimentally-induced spinal facet joint OA injury. As such our results strengthen our understanding of the physiologic expression of neurogenic inflammatory mechanisms in the clinical manifestation of joint inflammation. These findings may have potential implications to the pathophysiology of OA in that the proinflammatory effects of SP may directly influence cartilage and chondrocyte homeostasis, with potential implications to joint inflammation and OA.

SP is a key neurotransmitter released during neurogenic inflammation and important in the pathophysiology of OA.^4,35,36 SP is a pro-inflammatory neurotransmitter acting on the neurokinin-1 (NK-1) receptor to increase vascular permeability, promote vasodilation and evoke powerful proinflammatory responses within joints and tissues.^20,37,38 Levine was the first to propose that intraarticular SP concentration may be directly correlated to the degree of inflammation observed in OA, and that the origin of intraarticular SP may be neurogenic.^19,39 This has been shown experimentally in several animal studies where intraarticular injection of SP led to dose-dependent synovial plasma extravasation^20,40 and increased OA severity.^13,19

The existing body of research highlights causal associations between experimentally induced unilateral arthritis and the spread of inflammation and articular degeneration in contralateral, homologous joints. Acute onset inflammatory responses have been reported in the contralateral limb after injections of crystals (monosodium urate, calcium pyrophosphate dihydrate, hydroxyapatite, calcium oxalate) into the footpad⁴¹ and hypertonic saline into the hindpaw⁴² of rats. Experimental unilateral arthritis models of OA using complete Freund’s adjuvant (CFA) injections also exhibit heightened inflammatory responses with significant increases in cartilage degradation (42% decrease in proteoglycan synthesis)⁴ in a dose-dependent relationship¹¹ in the contralateral limb. A dose-response association has also been observed clinically where bilateral knee OA involvement has been more commonly observed with increased severity of arthritis in one knee joint.⁹ On the other hand, Decaris et al.⁴ showed a significant acute decreased in proteoglycan anabolism in the ipsilateral knee patella 6 hours after subcutaneous injection of CFA in the hind paw. This effect was minimized with a pretreatment with NK-1 antagonist thereby suggesting a neurogenic mechanism dependence.⁴

Although most notably expressed in neuronal cells, SP and the NK-1 receptor have been identified in a growing number of non-neuronal cells. For example, SP is expressed in immune cells, including monocytes, T lymphocytes, and eosinophils,^38,40 where it plays a role in regulating cytokine production (interleukin [IL]-1, IL-6, IL-8, tumor necrosis factor-α), inflammation and immune responses. SP expression has also been reported in the costal cartilage of mice²⁷ and in adult human chondrocytes,⁴³ where its precise role is still unclear. In this context, SP appears to play a key role in mechanotransduction as evidenced by its ability to enhance the electrophysiological response of human articular chondrocytes exposed to mechanical stimulation²⁷ through its direct action on the NK-1 receptor. In addition, studies in vitro have shown that exogenous SP stimulates chondrocyte proliferation suggesting a role in the maintenance of the articular structure.⁴³ Interestingly, our findings show that a basic expression of SP is presented in all groups studied supporting its potential function in the maintenance of the cartilage structure.

The knee articular cartilage is structurally composed of 3 important cartilage zones and their combination provides and important functional advantage absorbing and transmitting mechanical forces.³⁴ Curiously, chondrocytes have a dual role synthetizing key component of the cartilage matrix (collagen and proteoglycan) as well as producing enzymes that degrade matrix components (collagenases and proteinases).^6,27,34 It is widely described that the superficial zone is the smallest zone containing a relatively high number of flattened chondrocytes and its integrity is crucial to the protection and maintenance of deeper layer homeostasis.³⁴ Functionally, the middle and deep zones are responsible for most of the resistance to compressive forces.^5,34 Karahan et al.⁴⁴ demonstrated that after a series of physical exercise an increased expression of SP in the middle and deep cartilage zones becomes evident in dogs. Its upregulation was then suggested to positively contribute to cartilage matrix organization in response to mechanical stimulation.⁴⁴ In the present study chondrocytes located in the superficial cartilage zones had higher positive-SP chondrocyte density regardless of the group studied. The superficial layer in the control group presented approximately 35% higher density of SP-positive chondrocytes compared with middle and deep zones. The superficial layer in the spinal facet joint-OA surgery group is approximately 50% and 70% higher than the middle and deep zones, respectively. Yet, the sham intervention group displayed a 30% and 45% larger density at the superficial zone compared to middle and deep zones, although statistical significance was only reached between superficial and deep zones within this group. The interpretation of these findings is equivocal on the basis of the existing body of literature in this field. The significant difference of the overall SP chondrocyte expression as well as the significant difference in the superficial layer between spinal facet OA-injury to sham intervention (37%) and control (31%) groups could represent an adaptive response of chondrocytes to subsequently altered gait patterns and mechanical joint stresses. It has been previously shown that lumbar spine facet injury significantly affects gait biomechanics in rodents.⁴⁵ In this context, however, given the role of deeper cartilage zones in the resistance of compressive forces compared to superficial zone⁴⁶ and the positive influence of SP in cartilage mechanotransduction⁴⁷ and maintenance of matrix organization⁴⁴ postexercises, we would expect to see a larger density of SP immunoreactivity in deeper zones, which was not apparent. Due to the complexity of the influence of joint loading on cartilage and chondrocyte metabolism, SP may have a different response to the pathological context presented in our work compared with the physiological stress-force loading associated with exercise. Further studies are needed to understand the pathological versus physiological differences in the expression of SP in the cartilage.

In contrast, enhanced expression of SP in chondrocytes could also represent a maladaptive subclinical response to neurogenic responses within the knee as shown by the larger modified Mankin score in the present study. It has been shown that spinal facet joint injury leads to increased sensory inputs within the spinal cord segments leading to chronic pain and distal inflammation.^15,16,48 Existing animal models suggest that neurogenic inflammation is mediated by the dorsal root reflex (DRR), characterized by proximal (central) terminal depolarization of primary afferents (A-delta and C-fibers) and leading to antidromic release of neurotransmitters (i.e., SP and calcitonin gene related-peptide [CGRP]) into peripheral tissues.⁴⁹ Experimental evidence also suggests that the DRR is mediated through neurosegmental innervated pathways, both ipsilaterally and contralaterally.^15,49,50 Interestingly, peripheral A-delta and C-fibers terminals innervating knee joint tissues present central projections to lumbar segments 4, 5, and 6.⁵¹ Although consensus is that cartilage tissue is aneural,^34,46 nociceptive fibers have been previously reported within osteoarthritic joints and articular cartilage, introducing a potentially important pathway for the mediation of neurogenic inflammation in cartilaginous tissues.^52,53 CGRP producing nerve fibers have been identified in contact with surface chondrocytes of developing cartilaginous tissue.^27,53 In the context of OA disorder, Im et al.⁵⁴ elegantly demonstrated that IL-1 stimulates the expression (mRNA) and the protein secretion of the SP and its receptor NK1 in arthritic human adult articular chondrocytes. Their binding was thought to accelerate cartilage degradation by enhancing the production of cartilage-degrading enzymes (MMP-13) as well as attenuating proteoglycan deposition in adult human cartilage.⁵⁴ SP-mediating induction of MMP-13 was showed to occur via Raf-ERK MAPKs and NFκB intracellular pathways in human adult articular chondrocytes.⁵⁴ Similarly, cultured bovine chondrocytes with the C-terminal fragment of SP but not SP stimulated the production of prostaglandin, calcium and collagenase in chondrocytes. Since SP can be cleaved in the synovial fluid chondrocyte-mediated cartilage pathology may also be influenced by SP residual fragments.²⁹ Additionally, expression of the vascular endothelial growth factor (VEGF) as well as the co-expression of nerve growth factor (NGF) and CGRP was recently demonstrated in the superficial zone chondrocytes of the human medial tibial plateau OA cartilage but not in a non-arthritic cartilage.²⁶ Since these growth factors, especially NGF, is widely known to stimulate SP and CGRP synthesis, axonal transport, and release in arthritic joints^18,53 their participation in the superficial cartilage breakdown was suggested.²⁶ The superficial zone is in contact with synovial fluid, contains large density of chondrocytes and is responsible for most of the tensile properties of cartilage.³⁴ It has been shown that the superficial zone is more susceptible to chondrocyte apoptosis and chondrocyte catabolism.^34,55 Interestingly, the superficial zone chondrocytes, in the presence of synovium IL-1 was attributed to be more susceptible to catabolic and damage effects than deeper zones.^55,56 Thereby, given the close relationship that superficial chondrocytes present with innervated synovial tissue it is possible that the L4-L5 facet lesion triggered an antidromic transport of SP or another neuropeptide, which may have initiated catabolic changes in the synovial environment, resulting in the enhanced chondrocytes SP mainly in the superficial zone, early osteoarthritic changes and significantly elevated modified Mankin scores in the ipsilateral knee.

Our findings should be interpreted in light of the fact that the surgically induced primary pathology in our model was mild when assessed histologically and compared with that of Henry et al.,⁴⁸ which showed a completely degenerated facet joint. Previous research has demonstrated a robust dose-response relationship between the severity of primary pathology and intensity of contralateral neurogenic inflammatory responses.^9-11 Similarly, a strong inverse association has been reported between SP concentration and proteoglycan production in cartilage.⁵⁴ These observations point to the importance of future studies in advancing this line of inquiry by employing a more severe injury model. In addition, we did not control for nonsegmentally linked tissues in our design as we were primarily interested in the causal association of experimentally induced spinal injury and neurogenic responses within neurosegmentally linked heterologous tissues. Future research should investigate these effects within nonsegmentally related heterologous tissues as well as neurosegmentally linked contralateral heterologous tissues to provide greater insight into the segmental and nonsegmental (systemic) physiologic mechanisms potentially contributing to these observations. In addition, psychophysical outcomes would add additional insight into potential effects of altered gait and/or joint loading on SP expression.

While a growing body of literature reports increased expression of SP within the contralateral homologous joint after experimentally induced monoarthritis, our findings are the first to demonstrate similar increases in SP expression ipsilaterally within neurosegmentally linked heterologous cartilage tissue 28 days post–lumbar facet OA injury. Here, we report increased expression of SP within the chondrocytes of knee cartilage after surgically induced facet compression injury at the common L5-L6 spinal level. In addition, a likely enhancement of SP preference in the superficial cartilage zone was also showed at 28 days post–lumbar facet OA injury compared to sham and control which may have contributed to early OA development. These findings potentially support our hypothesis that pro-inflammatory molecules may spread via neurogenic mechanisms to neurosegmentally linked heterologous tissues (in addition to homologous tissues) and contributes important insight into the mechanisms facilitating the clinical manifestation of neurogenic inflammation. Future directions should advance the findings of this study by investigating the dose-response relationship between primary pathology and neurogenic inflammatory response within heterologous tissues such as synovial membrane and synovial fluid, as well as assessing histological and radiological changes within the affected knee joint. Given the aging societal demographic and growing burden of OA, this line of research is timely and urgent to informing our understanding of the pathophysiologic mechanisms and clinical manifestation of chronic inflammatory joint disease and OA.