In situ observation of hyaluronan molecular weight effectiveness within articular cartilage lubrication

Citation: Rebenda, D.; Ranuša, M.; ˇ Cípek, P.; Toropitsyn, E.; Vrbka, M. In Situ Observation of Hyaluronan Molecular Weight Effectiveness within Articular Cartilage Lubrication. Lubricants 2023,11, 12. https://doi.org/10.3390/ lubricants11010012 Received: 11 November 2022 Revised: 14 December 2022 Accepted: 24 December 2022 Published: 30 December 2022 Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). lubricants Article In Situ Observation of Hyaluronan Molecular Weight Effectiveness within Articular Cartilage Lubrication David Rebenda 1,2,* , Matúš Ranuša 1,* , Pavel ˇ Cípek 1, Evgeniy Toropitsyn 3,4 and Martin Vrbka 1 1Faculty of Mechanical Engineering, Brno University of Technology, Technická2896/2, 616 69 Brno, Czech Republic 2Footwear Research Centre, University Institute, Tomas Bata University in Zlin, Nad Ovˇcírnou IV 3685, 760 01 Zlín, Czech Republic 3Contipro a.s., DolníDobrouˇc 401, 561 02 DolníDobrouˇc, Czech Republic 4Biocev, First Faculty of Medicine, Charles University, Pr˚umyslová595, 252 50 Vestec, Czech Republic *Correspondence: [email protected] (D.R.); [email protected] (M.R.) Abstract: Degradation of the articular cartilage (AC) structure due to osteoarthritis significantly influences its friction and lubrication mechanisms. Injection with exogenous hyaluronic acid (HA) is one of the ways to slow down the progress of these changes. The present paper aims to determine the effect of HA on the friction and lubrication processes of the synovial joint model. The main emphasis is placed on the effect of HA molecular weight (MW) on the coefficient of friction (COF) and the interactions between HA and other constituents of synovial fluid (SF). Frictional measurements between the AC surface and the glass were performed with simultaneous in situ observation of the contact zone by fluorescence microscopy. Using this methodology, a decrease in AC COF with an increase in the fluorescence intensity emitted from contact with HA was observed, while the phenomenon was found to be MW-dependent. These findings demonstrate that high-MW HA is more effective within a resumption of healthy AC lubrication due to a better adhesion to the AC surface. Keywords: articular cartilage; friction; lubrication; hyaluronic acid; viscosupplementation 1. Introduction Healthy articular cartilage (AC) is a self-lubricating system which plays an important role in the life-time longevity of synovial joints. Under physiological conditions, surprisingly low values of the coefficient of friction (COF), as low as 10 −3 , at relatively high pressures were reported [ 1 , 2 ]. These excellent tribological properties are commonly attributed to adaptive multimode lubrication [ 3 ]. Due to this complex lubrication regime, AC superior lubricity is controlled by interstitial fluid pressurization [ 1 , 4 ] and by a boundary lubricating layer on the AC surface [ 5 ]. These boundary lubricating layers are mainly composed of a combination of hyaluronic acid (HA) [ 6 , 7 ], lubricin (PRG4) [ 8 , 9 ], phospholipids [10,11] and proteins [12]. Osteoarthritis (OA) is one of the most rapidly rising diseases among elderly people in developed countries. Lancet’s research [ 13 ] revealed that more than 303 million people in the world were diagnosed with osteoarthritis between 1990 and 2017. The OA joint exhibits a worn AC surface and, as a consequence, increased friction due to the breakdown in lubrication mechanism [ 14 ]. The composition of synovial fluid (SF) is changing [ 15 ], plus SF is diluted by an inflammatory effusion. Due to this, the production of HA by synoviocytes is reduced [ 16 ]. On the other hand, HA degradation is promoted by hyaluronidases [ 17 ] and reactive oxygen species [ 18 ]. Consequently, the concentration and molecular weight (MW) of HA in OA SF are significantly lower [ 19 ]. Since HA is the main constituent of SF that affects its rheological properties, the viscosity and viscoelastic properties of SF are aggravated [20]. Lubricants 2023,11, 12. https://doi.org/10.3390/lubricants11010012 https://www.mdpi.com/journal/lubricants Lubricants 2023,11, 12 2 of 15 To improve the rheological properties of the OA SF and restore lubrication mechanisms within OA AC fibrous structure, SF can be doped with endogenous HA. Despite the wide application of this treatment method, also known as viscosupplementation, its positive effects compared to the placebo are still the subject of debate [ 21 – 23 ]. The shear rate in the human joint can vary between 0.01 s −1 and 10 5 s −1 [ 24 ]. Therefore, a significant increase in SF viscosity by endogenous HA at such high shear rates, due to its shear-thinning behavior, does not seem to be the true mechanism of action. Because of that, the clarification of the HA role during the formation of a supramolecular boundary layer on the AC surface seems substantial for redeeming hydration lubrication within OA AC. HA, due to molecular entanglement or due to the hydrophobic/hydrophilic nature of the molecules, tends to bind to PRG4 on the AC surface [ 9 , 25 ] to form a highly viscous layer. However, it seems as though the HA itself does not support the extremely low friction of AC [26]. The formation of a HA layer seems to be a basis for an interaction with phospholipids to form a complex structure [ 27 , 28 ] that exposes highly hydrated phosphocholine head groups [ 29 ]. During the mutual movement of AC surfaces, hydrophilic headgroups of phospholipids exchange water molecules by diffusion. Due to this lubrication mechanism, the COF within the AC ranges in the thousands. The performance of HA within AC friction is MW-dependent. Kwiecinski et al. [ 30 ] reported an approximately linear dependence between HA MW and COF for simple HA solutions. Surprisingly, this effect was not observed for HA+PRG4 solutions. Liu et al. [ 26 ] reported a lower COF and a more stable lubricating film for high-MW HA than for low-MW HA. Shorter chains are weakly attached to the AC surface emulating layer. As a consequence, HA with the lubricating phospholipid layer is more easily removed by shear during sliding. Hydration lubrication seems to be essential for AC homeostasis. Despite AC degradation due osteoarthritis, residuals of hydration lubrication were observed also for phospholipids extracted from OA SF [ 31 ]. However, higher friction was reported as a result of a hemifusion of exposed phospholipid bilayers under physiological contact pressures. Slip plane shift from hydrophobic-tails/hydrophobic-tails to hydrophobic-tails/AC surfaces is connected with a higher dissipation energy caused by a braking of van de Waals bonds between molecules. Therefore, the prevention of hemifusion and the consequent shift of the slip plane could be beneficial for AC lubrication [ 32 ]. HA should enable the formation of phospholipid multilayers on the AC surface [33] and inhibit their degradation [34]. In this study, our objective was to investigate the effect of admixed HA on the COF in the cartilage-on-glass contact lubricated by artificial OA SF. To determine the mutual interactions between the SF constituents and directly observe their role within the formation of the AC boundary lubricating layer, fluorescence microscopy was utilized for the in situ observation of contact during the measurements. The main emphasis was placed on the role of HA MW. Therefore, three viscosupplement-simulating HA solutions with different MW were individually mixed with OA SF. Based on the literature, we know that interaction between HA and phospholipids plays an important role within AC lubrication. Apart from that, the interaction between HA and other SF may also play an important role within redeeming AC lubrication. According to Murakami et al. [ 35 ] or Yarimitsu et al. [ 36 ], positively charged molecules of protein γ -globulin may interact with negatively charged molecules of HA to create complex structures which contribute to AC friction. Therefore, we also focused on the effect of HA MW on the formation of a protein boundary layer on the AC surface. 2. Materials and Methods 2.1. Friction Apparature Measurements were performed using a custom-built reciprocating simulator, which was described in detail in a study by ˇ Cípek et al. [ 37 ]. The main benefit of this device is the simultaneous measurement of friction forces within the contact and in situ observation of the contact by fluorescence microscopy. Sliding tests were conducted in a pin-on-plate configuration, while COF and an intensity of the light emitted from the contact as a Lubricants 2023,11, 12 3 of 15 function of the time were investigated. The contact pair consisted of a movable glass plate and a stationary AC specimen which was mounted in a loading mechanism of the tribometer under the glass plate. The scheme of the whole friction apparatus and fluorescence microscope is shown in Figure 1. The AC sample was loaded with a constant load of 10 N against the glass plate, while the glass plate was undergoing a reciprocating sliding motion with a constant sliding speed of 10 mm/s and the stroke length was set to 20 mm. The contact pair was fully flooded with SF or a mixture of SF and HA, while the contact was heated to 37 ◦ C via heating cartridges mounted in a steel chamber with test fluid. The loading and frictional forces were monitored by two single-point load cells which were mounted on the loading mechanism of the tribometer. Lubricants 2023, 11, x FOR PEER REVIEW 3 of 15 2. Materials and Methods 2.1. Friction Apparature Measurements were performed using a custom-built reciprocating simulator, which was described in detail in a study by Čípek et al. [37]. The main benefit of this device is the simultaneous measurement of friction forces within the contact and in situ observation of the contact by fluorescence microscopy. Sliding tests were conducted in a pin-on-plate configuration, while COF and an intensity of the light emitted from the contact as a function of the time were investigated. The contact pair consisted of a movable glass plate and a stationary AC specimen which was mounted in a loading mechanism of the tribometer under the glass plate. The scheme of the whole friction apparatus and fluorescence microscope is shown in Figure 1. The AC sample was loaded with a constant load of 10 N against the glass plate, while the glass plate was undergoing a reciprocating sliding motion with a constant sliding speed of 10 mm/s and the stroke length was set to 20 mm. The contact pair was fully flooded with SF or a mixture of SF and HA, while the contact was heated to 37 °C via heating cartridges mounted in a steel chamber with test fluid. The loading and frictional forces were monitored by two single-point load cells which were mounted on the loading mechanism of the tribometer. Each experiment, i.e., the measurement with one AC sample and one type of lubricant, consisted of three loaded phases which were separated by two unloaded phases (Figure 1). During the loaded phases, which lasted 120 s, the AC was loaded against the glass plate and the glass plate was performing a reciprocating sliding motion. Therefore, the traveled sliding distance during one loaded phase is approximately 1100 mm. To investigate the effect of AC rehydration on friction and lubrication, AC samples were unloaded but immersed in the tested lubricant for approximately 300 s in between loaded phases. In addition to rehydration, this relatively long time was also necessary to download the data from the camera which recorded the course of the measurement. To ensure the relevance of the results, each experiment (3 loaded phases) with one type of lubricant was repeated 5 times with a fresh AC sample and fresh lubricant. Figure 1. Research plan. The optical method based on the principle of fluorescence microscopy was utilized for in situ observation of the contact area. For this purpose, an optical module with a Figure 1. Research plan. Each experiment, i.e., the measurement with one AC sample and one type of lubricant, consisted of three loaded phases which were separated by two unloaded phases (Figure 1). During the loaded phases, which lasted 120 s, the AC was loaded against the glass plate and the glass plate was performing a reciprocating sliding motion. Therefore, the traveled sliding distance during one loaded phase is approximately 1100 mm. To investigate the effect of AC rehydration on friction and lubrication, AC samples were unloaded but immersed in the tested lubricant for approximately 300 s in between loaded phases. In addition to rehydration, this relatively long time was also necessary to download the data from the camera which recorded the course of the measurement. To ensure the relevance of the results, each experiment (3 loaded phases) with one type of lubricant was repeated 5 times with a fresh AC sample and fresh lubricant. The optical method based on the principle of fluorescence microscopy was utilized for in situ observation of the contact area. For this purpose, an optical module with a fluorescence microscope was situated above the contact. The principle of fluorescence phenomena consists of three stages. Firstly, photons excited by the light source are absorbed by the fluorescent dye contained in the tested lubricant. Consequently, partial energy dissipation occurs to increase the fluorescence emission. Due to this energy dissipation, emitted photons exhibit a lower energy and the emitted radiation has a wavelength longer than excited radiation. Due to this shift, excitation and emission can be separated and the fluorescence yield can be determined. In our case, contact was observed using a light-emitting diode. A fluorescein isothiocyanate (FITC) or tetramethyl rhodamine (TRITC) filter was placed behind the light- Lubricants 2023,11, 12 4 of 15 emitting diode to achieve the required wavelength of emitted and excited light. The excitation wavelengths are 490 nm for the FITC filter and 557 nm for the TRITC filter. Emission wavelengths are 525 nm and 576 nm, respectively. Data of the in situ observation were recorded with a high-speed complementary metal-oxide semiconductor (CMOS) camera Andor Neo 5.5 (Andor, Belfast, UK). 2.2. Lubricants and Samples The contact was lubricated with a model SF which was mixed, during the later stages of the measurements, in a 1:1 ratio with different MW HA solutions. The exact composition of the model SF was based on the extensive analysis of SFs obtained from various groups of orthopedic patients by Galandákováet al. [ 15 ]. On the basis of these findings, a model SF whose composition should correspond to the SF of patients with osteoarthritis was designed. The exact composition was as follows: albumin = 24.9 mg/mL , γ-globulin = 6.1 mg/mL , HA = 1.49 mg/mL, phospholipids = 0.34 mg/mL. Phosphate-buffered saline (PBS) was used as a basic solution in which the individual components were dissolved. The following products were used for the preparation: bovine serum albumin (powder, ≥ 96%; A2153, Sigma-Aldrich, St. Louis, MO, USA), γ -globulin from bovine blood (powder, ≥ 99%; G5009, Sigma-Aldrich, St. Louis, MO, USA), sodium hyaluronate HySilk (powder, quality class— cosmetic; molecular weight = 820–1020 kDa, Contipro, DolníDobrouˇc, Czech Republic) and Lα -phosphatidylcholine (powder, Type XVI-E, lyophilized powder; ≥ 99%; vesicles form; P3556, Sigma-Aldrich, St-Louis, MO, USA). The main focus of this study was the role of the MW of HA which is added to the OA synovial joint during viscosupplementation. Therefore, OA SF was mixed with three different HA solutions which contained HA with a MW of 500 kDa (HA500), 1000 kDa (HA1000) or 1600 kDa (HA1600). In all three cases, the concentration of the HA solution was the same: 20 mg/mL. The concentration and MW of HA was based on commercial viscosupplements which were analyzed in our previous study [ 38 ]. Solutions were prepared from HA powder (Contipro, DolníDobrouˇc, Czech Republic) by the dissolution of a required amount of powder in PBS. When using fluorescence microscopy, it is possible to focus on the role of the individual SF constituent. To do so, the selected component (albumin, γ -globulin, HA) had to be labeled with a fluorescent dye. Albumin was doped with rhodamine-B-isothiocyanate (283,924, Sigma-Aldrich, St. Louis, MO, USA) and γ -globulin or HA was doped with fluorescein-isothiocyanate (F7250, Sigma-Aldrich, St. Louis, MO, USA). In our case, the fluorescence microscope is able to observe only one specific constituent during measurement. Therefore, experiments with the same lubricant were repeated multiple times, while one of the previously mentioned components was observed. In the case of clear SF, only albumin and γ -globulin were analyzed. For mixtures of SF and HA solutions, both proteins and admixed HA were labeled and analyzed. For better illustration, the overview of tested lubricants and marked components is specified in detail in Table 1. Intact AC specimens with underlying subchondral bone were removed from the bovine femur, tibia or patella. Whole bovine stifle joints with surrounding tissues were received from a local slaughterhouse on the day of slaughter. Cylindrical samples with an inner diameter of 9.7 mm were extracted by a custom-made ejector. The extracted samples were visually inspected to exclude the samples with damaged AC tissue or frayed edges, whereas these circumstances could potentially influence the results of the measurements. The extracted samples were deeply frozen at − 20 ◦ C in PBS to slow down the degradation of the tissue. Prior to the measurements, frozen AC samples were removed from the freezer and thawed naturally at laboratory temperature. As a counter face to the bovine AC, optically smooth glass B270 was selected. Lubricants 2023,11, 12 5 of 15 Table 1. Overview of the used lubricants. Fluorescently labeled components of SF are highlighted. No. Albumin γGlobulin HA Phospholipids Viscosupplement Fluorescently labeled albumin 124.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL - 224.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA500—20 mg/mL 324.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA1000—20 mg/mL 424.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA1600—20 mg/mL Fluorescently labeled γ-globulin 5 24.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL - 6 24.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA500—20 mg/mL 7 24.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA1000—20 mg/mL 8 24.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA1600—20 mg/mL Fluorescently labeled viscosupplement 9 24.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA500—20 mg/mL 10 24.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA1000—20 mg/mL 11 24.9 mg/mL 6.1 mg/mL 1.49 mg/mL 0.34 mg/mL HA1600—20 mg/mL 2.3. Data Processing and Statistical Analysis During measurements, the AC pin was mounted in a pin holder which was connected to two single-point load cells. Based on the vales of loading and friction forces recorded by these load cells, the COF as a function of time was analyzed. All frictional measurements (one type of lubricant) were conducted five times with fresh AC sample. Moreover, we assumed that fluorescent dyes do not affect the COF values. Therefore, average values and standard errors of means (SEM) from ten or fifteen measurements were calculated and will be presented further. From the CMOS camera data, the average fluorescence intensity emitted from contact was analyzed. Fluorescence intensity can be considered as a dimensionless film thickness since a direct relationship between fluorescence intensity and film thickness measurement for SF lubricated contact was already published by Neˇcas et al. [ 39 ]. However, due to an AC porous structure which is connected with lubricant penetration into deeper layers of the structure [ 40 , 41 ], a calibration curve cannot be obtained. Following this inconvenience, results can be analyzed only in a qualitative manner, i.e., by analyzing the trends of film thickness development and influence of SF composition. In addition to fluorescence intensity, the images recorded with the CMOS camera were processed by custom designed software, which was described in detail in a study by ˇ Cípek et al. [ 42 ]. This software, based on the principle of image segmentation, calculates the particle count and the average size of particles in the contact area. Based on these values, percentage coverage of the contact area by adsorbed proteins was analyzed. Adsorbed particles of HA were not detected. Therefore, these data are not presented. Due to repeated measurements, the data will also be presented as mean ±SEM. A one-way analysis of variance (ANOVA) and Tukey’s post hoc test was used to compare the COF, the fluorescence intensity or the percentage cover of the contact area for various tested lubricants. In all ANOVA analyses, α was set to 0.05 and results with p< 0.05 were deemed as statistically significant. All ANOVA analyses were performed using TIBCO Statistica®13.5.0 (TIBCO Software Inc., Palo Alto, CA, USA). 3. Results 3.1. Frictional Measurements The results of COF as a function of time for all tested lubricants are shown in Figure 2. As can be seen, independently of the tested lubricant, the COF continuously increases during all measurement sub-steps, which is mainly attributed to the biphasic structure of the AC. After 120 and 240 s, a rapid decrease in COF can be observed for all tested lubricants. This is attributed to the rehydration of the AC structure by the unloading phases mentioned Lubricants 2023,11, 12 6 of 15 above, which separated every two measurement sub-steps. The highest COF was measured for pure SF. For this lubricant, the most significant increase in COF was also observed during measurements. In the case of the last measurement sub-step, a gradual increase in COF from 0.061 ± 0.0086 (mean ± SEM) to 0.121 ± 0.0109 was observed. The COF of pure SF was reduced by adding HA in a MW-dependent manner throughout the tested range, i.e., 500–1600 kDa, and the absolute increase in COF during measurements was declining. For the case of HA500, the COF increased from 0.036 ±0.0095 to 0.074 ±0.0151 during the last measurement. HA1000 exhibited slightly lower COF values. During the same measurement sub-step, COF increased from 0.029 ±0.0045 to 0.058 ±0.0088 . Compared to HA500, HA1600 exhibited half the COF values. An increase from 0.019 ±0.0035 to 0.039 ±0.0057 was observed during the last measurement sub-step. Lubricants 2023, 11, x FOR PEER REVIEW 6 of 15 3. Results 3.1. Frictional Measurements The results of COF as a function of time for all tested lubricants are shown in Figure 2. As can be seen, independently of the tested lubricant, the COF continuously increases during all measurement sub-steps, which is mainly attributed to the biphasic structure of the AC. After 120 and 240 s, a rapid decrease in COF can be observed for all tested lubricants. This is attributed to the rehydration of the AC structure by the unloading phases mentioned above, which separated every two measurement sub-steps. The highest COF was measured for pure SF. For this lubricant, the most significant increase in COF was also observed during measurements. In the case of the last measurement sub-step, a gradual increase in COF from 0.061 ± 0.0086 (mean ± SEM) to 0.121 ± 0.0109 was observed. The COF of pure SF was reduced by adding HA in a MW-dependent manner throughout the tested range, i.e., 500–1600 kDa, and the absolute increase in COF during measurements was declining. For the case of HA500, the COF increased from 0.036 ± 0.0095 to 0.074 ± 0.0151 during the last measurement. HA1000 exhibited slightly lower COF values. During the same measurement sub-step, COF increased from 0.029 ± 0.0045 to 0.058 ± 0.0088. Compared to HA500, HA1600 exhibited half the COF values. An increase from 0.019 ± 0.0035 to 0.039 ± 0.0057 was observed during the last measurement sub-step. For the end value of COF, ANOVA and Tukey’s post hoc test reported a statistically significant difference between the lubricants tested (p < 0.0001). The addition of all three HA samples caused a significant decrease in COF compared to pure SF (p = 0.009, p = 0.0003 and p = 0.0002). The COF values for HA500 were also significantly higher compared to HA1600 (p = 0.009). However, the differences between HA500 and HA1000 or HA1000 and HA1600 were not statistically significant (p = 0.46 and p = 0.27). Figure 2. COF as a function of a time for clear model SF and mixtures with different molecular weight HA solutions. Data are presented as mean ± SEM. 3.2. In Situ Observation of the Contact Area Together with frictional measurements, in situ observation of the contact area by fluorescence microscopy was performed. Figure 3A summarizes the data for all four lubricants for the case when the protein albumin was fluorescently labeled. The fluorescence intensity of the emitted light as a function of a time is presented. Figure 3B shows typical images of the contact area throughout the experiments for all four lubricants. For all tested lubricants, an increase in the fluorescence intensity after rehydration can be Figure 2. COF as a function of a time for clear model SF and mixtures with different molecular weight HA solutions. Data are presented as mean ±SEM. For the end value of COF, ANOVA and Tukey’s post hoc test reported a statistically significant difference between the lubricants tested (p< 0.0001). The addition of all three HA samples caused a significant decrease in COF compared to pure SF (p= 0.009, p= 0.0003 and p= 0.0002). The COF values for HA500 were also significantly higher compared to HA1600 (p= 0.009). However, the differences between HA500 and HA1000 or HA1000 and HA1600 were not statistically significant (p= 0.46 and p= 0.27). 3.2. In Situ Observation of the Contact Area Together with frictional measurements, in situ observation of the contact area by fluorescence microscopy was performed. Figure 3A summarizes the data for all four lubricants for the case when the protein albumin was fluorescently labeled. The fluorescence intensity of the emitted light as a function of a time is presented. Figure 3B shows typical images of the contact area throughout the experiments for all four lubricants. For all tested lubricants, an increase in the fluorescence intensity after rehydration can be observed, while the values of the fluorescence intensity during the measurement sub-steps are nearly constant. As in the case of frictional measurements, the highest values of fluorescence intensity were measured for clear SF. Throughout the experiment, the fluorescence intensity increased from (mean ± SEM) 10,812 ± 1134 to 20,673 ± 774. Unlike friction measurements, fluorescence intensity was not decreased by the addition of HA in a MW-dependent manner. Of the three mixtures, the highest fluorescence intensity values were obtained for a SF+HA1000 lubricant. For the remaining two lubricants, lower fluorescence intensity values were measured for the mixture of SF and HA1600. However, it should be mentioned Lubricants 2023,11, 12 7 of 15 that the fluorescence intensity of the lubricant SF+HA1000 and SF+HA1600 were quite similar during the first measurement sub-step. The fluorescence intensity values at the end of the measurement were 12,683 ± 331 (SF+HA500), 14,736 ± 666 (SF+HA1000), and 10,631 ±1289 (SF+HA1600). Lubricants 2023, 11, x FOR PEER REVIEW 7 of 15 observed, while the values of the fluorescence intensity during the measurement sub-steps are nearly constant. As in the case of frictional measurements, the highest values of fluorescence intensity were measured for clear SF. Throughout the experiment, the fluorescence intensity increased from (mean ± SEM) 10,812 ± 1134 to 20,673 ± 774. Unlike friction measurements, fluorescence intensity was not decreased by the addition of HA in a MW-dependent manner. Of the three mixtures, the highest fluorescence intensity values were obtained for a SF+HA1000 lubricant. For the remaining two lubricants, lower fluorescence intensity values were measured for the mixture of SF and HA1600. However, it should be mentioned that the fluorescence intensity of the lubricant SF+HA1000 and SF+HA1600 were quite similar during the first measurement sub-step. The fluorescence intensity values at the end of the measurement were 12,683 ± 331 (SF+HA500), 14,736 ± 666 (SF+HA1000), and 10,631 ± 1289 (SF+HA1600). For the end value of fluorescence intensity, ANOVA reported a statistically significant difference between the lubricants tested (p < 0.0001). Tukey’s post hoc test revealed a significant decrease in fluorescence intensity after the addition of HA to the SF for all HA samples tested: p = 0.0003 for SF+HA500, p = 0.002 for SF+HA1000, and p = 0.0002 for SF+HA1600. Significant differences were also observed between SF+HA1000 and SF+HA1600 (p = 0.032). Other differences, i.e., between SF+HA500 and SF+HA1000 (p = 0.43) or between SF+HA500 and SF+HA1600 (p = 0.44), were not statistically significant. Figure 3. (A) Albumin-based fluorescence intensity as a function of a time for clear SF and mixtures with different molecular weight HA solutions. (B) Fluorescence images of AC surfaces during measurements (the symbols in the top left corner refer to same line in the graph). Data are presented as mean ± SEM. The results of the fluorescence intensity measurements for the case of fluorescently labeled γ-globulin are shown in Figure 4A and typical images of the contact zone are shown in Figure 4B. As can be seen from the measurement results, a slight decrease in fluorescence intensity was measured during the measurement sub-steps for all the tested lubricants. For three of the four tested lubricants, an increase in fluorescence intensity was observed between the measurement sub-steps. Only for the SF+HA1600 mixture are the results of all three measurement sub-steps quite similar. In the beginning of measurements, the fluorescence intensity varies between 1510 and 1560. At the end of the measurements, the fluorescence intensity gradually decreases to a value between 1320 and 1410. This time, the highest values of fluorescence intensity were not measured for clear SF. During the first measurement sub-step, clear SF even reported the lowest fluoFigure 3. ( A ) Albumin-based fluorescence intensity as a function of a time for clear SF and mixtures with different molecular weight HA solutions. ( B ) Fluorescence images of AC surfaces during measurements (the symbols in the top left corner refer to same line in the graph). Data are presented as mean ±SEM. For the end value of fluorescence intensity, ANOVA reported a statistically significant difference between the lubricants tested (p< 0.0001). Tukey’s post hoc test revealed a significant decrease in fluorescence intensity after the addition of HA to the SF for all HA samples tested: p= 0.0003 for SF+HA500, p= 0.002 for SF+HA1000, and p= 0.0002 for SF+HA1600. Significant differences were also observed between SF+HA1000 and SF+HA1600 (p= 0.032 ). Other differences, i.e., between SF+HA500 and SF+HA1000 ( p= 0.43 ) or between SF+HA500 and SF+HA1600 (p= 0.44), were not statistically significant. The results of the fluorescence intensity measurements for the case of fluorescently labeled γ -globulin are shown in Figure 4A and typical images of the contact zone are shown in Figure 4B. As can be seen from the measurement results, a slight decrease in fluorescence intensity was measured during the measurement sub-steps for all the tested lubricants. For three of the four tested lubricants, an increase in fluorescence intensity was observed between the measurement sub-steps. Only for the SF+HA1600 mixture are the results of all three measurement sub-steps quite similar. In the beginning of measurements, the fluorescence intensity varies between 1510 and 1560. At the end of the measurements, the fluorescence intensity gradually decreases to a value between 1320 and 1410. This time, the highest values of fluorescence intensity were not measured for clear SF. During the first measurement sub-step, clear SF even reported the lowest fluorescence intensity values of all samples tested. The addition of HA1600 to SF resulted in an increase in fluorescence intensity. This mixture reported the highest fluorescence intensity values for all lubricants tested with labeled γ -globulin. The addition of HA500 and HA1600 to SF led to an increase in fluorescence intensity during the first measurement sub-step. However, during the second and the third measurement sub-steps, the fluorescence intensity was higher compared to the pure SF. For this set of experiments, the results of the HA mixtures were MW-dependent just for the third measurement sub-step. Lubricants 2023,11, 12 8 of 15 Lubricants 2023, 11, x FOR PEER REVIEW 8 of 15 rescence intensity values of all samples tested. The addition of HA1600 to SF resulted in an increase in fluorescence intensity. This mixture reported the highest fluorescence intensity values for all lubricants tested with labeled γ-globulin. The addition of HA500 and HA1600 to SF led to an increase in fluorescence intensity during the first measurement sub-step. However, during the second and the third measurement sub-steps, the fluorescence intensity was higher compared to the pure SF. For this set of experiments, the results of the HA mixtures were MW-dependent just for the third measurement sub-step. In general, the results of all the tested lubricants were quite similar. ANOVA did not reveal significant differences between lubricants (p = 0.39). The result of Tukey’s post hoc test also did not reveal statistically significant differences between any two lubricants. P values ranged between 0.66 and 0.96. Figure 4. (A) γ-Globulin-based fluorescence intensity as a function of a time for clear SF and mixtures with different molecular weight HA solutions. (B) Fluorescence images of AC surfaces during measurements (the symbols in the top left corner refer to same line in the graph). Data are presented as mean ± SEM. The last set of measurements was conducted for a mixture of pure SF and HA solutions with various MWs while the HA was labeled with a fluorescent dye. The results of fluorescence intensity are shown in Figure 5A and representative images of the contact area are shown in Figure 5B. Since we focused on the role of admixed HA, data for pure SF are missing in this set of results. For all three tested solutions, the fluorescence intensity decreased during the measurement sub-steps and increased in between them. The results of the individual mixture clearly depended on the MW of admixed HA. The highest values of fluorescence intensity were measured for the SF+HA1600 mixture, while the lowest values of fluorescence intensity, apart from the running-in phase during the first sub-step, were measured for the SF+HA500 mixture. At the end of the measurement, the fluorescence intensity values were as follows (mean ± SEM): 28 297 ± 910 (SF+HA500), 32 564 ± 1442 (SF+HA1000) and 36 031 ± 2231 (SF+HA1600). Although a fairly clear trend was observed in the increase in fluorescence intensity with increasing HA MW, ANOVA did not evaluate the differences between them as statistically significant (p = 0.168). Subsequently, Tukey’s post hoc also reported no statistically significant differences between SF+HA500 and SF+HA1000 (p = 0.19), SF+HA500 and SF+HA1600 (p = 0.264) or between SF+HA1000 and SF+HA1600 (p = 0.97). Figure 4. ( A ) γ -Globulin-based fluorescence intensity as a function of a time for clear SF and mixtures with different molecular weight HA solutions. ( B ) Fluorescence images of AC surfaces during measurements (the symbols in the top left corner refer to same line in the graph). Data are presented as mean ±SEM. In general, the results of all the tested lubricants were quite similar. ANOVA did not reveal significant differences between lubricants (p= 0.39). The result of Tukey’s post hoc test also did not reveal statistically significant differences between any two lubricants. pvalues ranged between 0.66 and 0.96. The last set of measurements was conducted for a mixture of pure SF and HA solutions with various MWs while the HA was labeled with a fluorescent dye. The results of fluorescence intensity are shown in Figure 5A and representative images of the contact area are shown in Figure 5B. Since we focused on the role of admixed HA, data for pure SF are missing in this set of results. For all three tested solutions, the fluorescence intensity decreased during the measurement sub-steps and increased in between them. The results of the individual mixture clearly depended on the MW of admixed HA. The highest values of fluorescence intensity were measured for the SF+HA1600 mixture, while the lowest values of fluorescence intensity, apart from the running-in phase during the first sub-step, were measured for the SF+HA500 mixture. At the end of the measurement, the fluorescence intensity values were as follows (mean ± SEM): 28 297 ± 910 (SF+HA500), 32 564 ± 1442 (SF+HA1000) and 36 031 ±2231 (SF+HA1600). Although a fairly clear trend was observed in the increase in fluorescence intensity with increasing HA MW, ANOVA did not evaluate the differences between them as statistically significant (p= 0.168). Subsequently, Tukey’s post hoc also reported no statistically significant differences between SF+HA500 and SF+HA1000 (p= 0.19), SF+HA500 and SF+HA1600 (p= 0.264) or between SF+HA1000 and SF+HA1600 (p= 0.97). 3.3. Coverage of Contact Area by Protein Clusters Apart from the average fluorescence intensity emitted from the contact area, the percentage cover of contact area by the protein-adsorbed clusters was evaluated from camera snaps as well. To do so, previously mentioned software based on image segmentation was utilized. The percentage of contact coverage by protein albumin is shown in Figure 6A and for γ -globulin in Figure 6B. For albumin, the results of the measurements with all the tested lubricants exhibited an increasing trend during the measurements. This clearly refers to the formation of a boundary lubricating layer on the AC surface while albumin is one of its components. For clear SF, albumin covered nearly 16% of the contact area at the Lubricants 2023,11, 12 9 of 15 end of the measurements. The addition of HA to SF partially inhibited the formation of albumin clusters. For all HA solutions, the percentage coverage of the contact area was lower compared to the pure SF. In addition, the results of all three HA solutions were quite similar. Lubricants 2023, 11, x FOR PEER REVIEW 9 of 15 Figure 5. (A) HA-based fluorescence intensity as a function of a time for mixtures of SF with different molecular weight HA solutions. (B) Fluorescence images of AC surfaces during measurements (the symbols in the top left corner refer to same line in the graph). Data are presented as mean ± SEM. 3.3. Coverage of Contact Area by Protein Clusters Apart from the average fluorescence intensity emitted from the contact area, the percentage cover of contact area by the protein-adsorbed clusters was evaluated from camera snaps as well. To do so, previously mentioned software based on image segmentation was utilized. The percentage of contact coverage by protein albumin is shown in Figure 6A and for γ-globulin in Figure 6B. For albumin, the results of the measurements with all the tested lubricants exhibited an increasing trend during the measurements. This clearly refers to the formation of a boundary lubricating layer on the AC surface while albumin is one of its components. For clear SF, albumin covered nearly 16% of the contact area at the end of the measurements. The addition of HA to SF partially inhibited the formation of albumin clusters. For all HA solutions, the percentage coverage of the contact area was lower compared to the pure SF. In addition, the results of all three HA solutions were quite similar. The percentage cover of contact area by albumin decreases to a value between 6.3 and 7.9%. Tukey’s post hoc test did not reveal statistical differences for end measurement values between any two HA mixtures. The p values ranged from 0.95 to 0.99. The decrease in percentage cover by HA addition for all three HA solutions was also considered statistically insignificant. Figure 5. ( A ) HA-based fluorescence intensity as a function of a time for mixtures of SF with different molecular weight HA solutions. ( B ) Fluorescence images of AC surfaces during measurements (the symbols in the top left corner refer to same line in the graph). Data are presented as mean ±SEM. Lubricants 2023, 11, x FOR PEER REVIEW 10 of 15 Figure 6. End values of percentage area of contact covered by protein (A) albumin and (B) γ-globulin before the first rehydration, before the second rehydration and at the end of measurements. Data are presented as mean ± SEM. In the case of γ-globulin, overall mean values of contact area coverage are considerably lower. End values of percentage cover just between 0.15 and 0.28% were reported. This indicates a considerably lower contribution of γ-globulin to the formation of a boundary lubricating layer on the AC surface. Furthermore, no significant differences were observed between the individual lubricants (p = 0.56). Therefore, HA and HA MW do not influence the formation of γ-globulin clusters within the AC-glass contact. Only a slight increase in percentage coverage during measurement was reported for the SF+HA1000 mixture and a slight decrease for SF+HA1600. 4. Discussion 4.1. General Discussion The present study introduces the frictional behavior and provides insight into AC lubrication by a fluorescence microscopy apparatus. The main emphasis was placed on the role of HA MW with which SF is doped during the viscosupplementation of the OA synovial joint. The results were confronted with pure OA SF, while some significant effects on AC friction and lubrication were observed. Statistical analysis revealed a significant decrease in COF after the addition of all HA solutions to pure OA SF, as well as significant differences in the results between HA500 and HA1600. Furthermore, a simple regression analysis on the end values of the COF measurements (Figure 7A) quite clearly revealed a linear dependency between the COF and the MW of admixed HA. In all three cases, the coefficient of determination varied between 92.1 and 93.8%. A strong dependence between COF and HA MW was also reported by Kwiecinski et al. [30]; however, a linear dependency between COF and log HA MW was reported. This difference is trends may be caused by the tested lubricants. In their case, simple HA solutions of HA mixed with PRG4 were tested. However, for the HA+PRG4 mixtures, the results were not MW-dependent. Figure 6. End values of percentage area of contact covered by protein ( A ) albumin and ( B ) γ -globulin before the first rehydration, before the second rehydration and at the end of measurements. Data are presented as mean ±SEM. The percentage cover of contact area by albumin decreases to a value between 6.3 and 7.9%. Tukey’s post hoc test did not reveal statistical differences for end measurement values between any two HA mixtures. The pvalues ranged from 0.95 to 0.99. The decrease in percentage cover by HA addition for all three HA solutions was also considered statistically insignificant. In the case of γ -globulin, overall mean values of contact area coverage are considerably lower. End values of percentage cover just between 0.15 and 0.28% were reported. This indicates a considerably lower contribution of γ -globulin to the formation of a boundary lubricating layer on the AC surface. Furthermore, no significant differences were observed between the individual lubricants (p= 0.56). Therefore, HA and HA MW do not influence the formation of γ -globulin clusters within the AC-glass contact. Only a slight increase in