Synovial joint organ system lubrication is a complex process influenced by mechanical, molecular, and fluid factors toward wear reduction and tissue homeostasis. In healthy systems, an effective lubricating molecular layer is formed between juxtaposed articular cartilage surfaces. Both structural and compositional changes in this layer can lead to aberrant lubrication, ineffective perturbation mitigation, and progressive system dysfunction. Because surgical advances have enabled the non-palliative repair of damaged cartilage surfaces without producing iatrogenic collateral tissue necrosis, an interest in normalizing this molecular lubricating layer has emerged as the therapeutic discipline of tribological surgical adjuvants [1
]. Yet, a confounding question lingers: can or should this molecular layer be normalized to enhance tissue rescue while pathologic phenotypes may persist post-treatment [2
]? Though knowledge of osteoarthritis’ phenotype spectrum appears ever increasing [3
], this study is designed as a framework to evaluate synovial fluid conditions that adjuvant candidates may encounter during surgical convalescence.
Hyaluronic acid (HA) and phospholipids (PL) are key constituents of the lubricating molecular layer in synovial joint organ systems. In addition, due to wide biocompatibility profiles and many established therapeutic uses, these molecules have been considered important candidates for tribological surgical adjuvant design [1
]. Both have been used clinically for short-term palliative relief, but have been ineffective as tribological agents when administered without surgical lesion repair first. As depicted in Figure 1
and Figure 2
, changes in these molecules occur commensurate with the degree of surface damage associated with system disease. Because damaged articular cartilage cannot support the formation of this lubricating molecular layer, these molecules form divergent complexes at and around the damage site rather than effective tribological structures. Fortunately, surgical repair has enabled the ability to create a suitable substrate upon which a surface-active phospholipid layer can reconstitute [4
]. However, the durability of this layer, and likewise the fate of surgical adjuvants delivered to improve the nature of this layer, remain at the mercy of osteoarthritic phenotypes which may persist and yield competing intra-articular processes. For this reason, tribological surgical adjuvant design must reconcile post-delivery system conditions.
Synovial fluid pH changes have been used to characterize relative degrees of inflammatory and non-inflammatory pathogenesis as an early effort toward treatment palliation [8
]; nevertheless, pH-dependent changes in the interaction between HA and PL have been afforded little clinical attention despite the profound impact on overall tribological function. In osteoarthritis, synovial fluid pH tends toward higher ranges [11
]. Addtionally, at the same time, HA concentration decreases and PL concentration increases when compared to normal synovial fluid profiles [12
]. More importantly, however, is that the species profiles within these molecular classes change qualitatively and quantitatively. For instance, HA chain length polydispersity is altered toward smaller-sized chains which display dissimilar intra- and inter-chain crosslinking affinities [13
]; and, PL composition is significantly transformed [14
] toward molecular saturation levels and fatty acid chain lengths less characteristic of those suited for synovial joint organ system lubrication [14
Previous model system observations indicate that lubrication-related molecules can be both chemically and physically attached to surfaces such that chemically grafted constructs provide better wear protection and show a decrease in frictional coefficients than do constructs attached only by physical mechanisms [16
]. Such observations support current in vivo findings that to produce surfaces which display durable wear properties, damaged articular cartilage requires physiochemical
surgical repair [4
]. Recognizing that the interfacial functional dynamics of the superficial active phospholipid layer display features of a sacrificial layer in dissipating perturbation forces and that favorable absorption–desorption kinetics are required to preserve that function [4
], intra-articular conditions post-delivery are important to study before exploring the tribological effects of various adjuvant preparations in limiting articular cartilage wear [17
In osteoarthritis, the total concentration of hyaluronic acid is slightly changed, but not all hyaluronic acid molecules are present in a long form due to weight distribution changes. The purpose of the present research is to investigate how the short faction of HA molecules behave in the event of an increased phospholipid concentration characteristic of osteoarthritis and to demonstrate that short HA molecular chains, by the interactions with PL, do not participate in the crosslinking of long HA molecular chains. This study examines the influence of pH on low molecular weight HA (50 nm; 20 kDa) and interactions amongst dipalmitoylphosphatidylcholine (DPPC) molecules, with pH = 7 representing healthy synovial fluid and pH = 8 representing osteoarthritic pathogenesis. Synovial fluid HA molecular weight ranges from ~4 kDa to ~8 MDa, wherein higher molecular weight species, which are selectively lost commensurate with osteoarthritis [19
], are associated with lubricating properties. Thus, examining low molecular weight HA is deemed to represent an osteoarthritic phenotype expected to display suboptimal features as an adjuvant candidate. DPPC is a saturated zwitterionic surfactant normally present in synovial joint organ systems at ~8–11% of total phosphatidylcholine species and which has been used clinically to treat osteoarthritis [15
Three DPPC concentrations are examined based upon normal concentration values (HA ~2.2 mg/mL; DPPC ~314.2 nmol/mL) expressed as a mass ratio of PL to HA, with
corresponding to healthy synovial joint organ systems,
to early osteoarthritis, and
to advanced osteoarthritis. The radius of gyration (Rg
) is evaluated to reflect relationships between protein size, folding rate, and spatial structure [21
]; root-mean-square deviation (RMSD), to analyze atomic position; and mean squared displacement (MSD), to assess the deviation time between molecular positions in an aqueous sodium chloride environment without other molecules. The Rouse model is applied to end-to-end vector analysis to evaluate polymer chain spread in space; and molecular surface area availability is evaluated with hydrophobicity and hydrogen bond number and their corresponding bond energies.
The relationships that exist between tribology and the mechanisms of synovial joint organ system lubrication, and hence surface degeneration, have been well-established [24
]; and, the restoration of surface properties remains an important therapeutic goal for biotribology now that surgical advances have enabled the repair of articular cartilage beyond palliative care to a suitable substrate upon which the surface-active phospholipid layer can reconstitute [4
]. This study begins to evaluate, as a prerequisite to tribological surgical adjuvant clinical testing, whether the lubricating molecular layer between juxtaposed articular cartilage surfaces can be normalized to further enhance tissue rescue despite osteoarthritic phenotypes that may persist after articular cartilage lesion repair and during surgical convalescence.
Based upon prior work regarding nanoscale friction in articulating systems [26
], this study further demonstrates that the nature of HA and PL interactions plays an important role in system efficiency; and, the synovial fluid changes in pH, hyaluronic acid polydispersity, and phospholipid concentration associated with osteoarthritis, individually and collectively, contribute to overall system dysfunction. Our results indicate that effective lubrication cannot be accomplished with a preponderance of intra-articular low molecular weight HA due to its inability to crosslink into effective supramolecular structures that are necessary for facilitated lubrication. Low molecular weight HA tends toward the formation of random coils, does not interlink, and produces lower non-strain-rate-dependent viscosity. Because molecular weight influences overall synovial fluid viscosity as described by the Mark–Houwink–Sakurada law [22
], three HA viscosity regimes undergoing Rouse and/or Zimm biopolymer chain dynamics have been described, determining that a molecular weight between 0.1 and 1.0 MDa optimizes crosslinking and network formation [34
]. This network formation appears based on the longitudinal diffusion coefficient being twice the transversal counterpart, which results in the same network spring constant ratios and a molecular length dependent on the rotational diffusion coefficient [34
Previous experimental data support our findings that HA Rg
is pH-dependent and its maximum Rg
value is observed at pH = 7 [35
]. We further demonstrate that low molecular weight HA molecules exhibit linear unfolded forms which are most hydrophobic (see Figure 13
), present the greatest surface accessibility (see Figure 10
), maximize crosslinking affinities, and display subdiffusion behavior and greater randomness at pH = 7.0 as presented in Figure 6
. Such findings would be expected under healthy synovial fluid conditions. Under osteoarthritic phenotypes, as represented by pH = 8, these features accede to a more compact folded form yielding decreased hydrophobicity, surface accessibility, and crosslinking affinities, as well as a more regular diffusion behavior and less randomness. This functional deterioration is in line with the trend toward lower molecular weight HA forms being associated with osteoarthritic synovial fluid, forms which have not been shown to play a significant role in facilitated lubrication regimes [13
]. Nonetheless, HA crosslinking mechanisms provide carrier functions, enable wetting, and maintain superficial active phospholipid layers important for lubrication [36
], and so crosslinking should remain an adjuvant design parameter.
In the presence of DPPC molecules, our data indicate a time delay in HA Rg
stabilization, reflecting the kinetic folding rate and less random behavior governed by the hydrophobic interactions between molecular species. A progressive decrease in Rg
with an increasing DPPC (Figure 3
and Figure 4
) concentration for pH = 7 was observed, while pH = 8 displayed irregular Rg
contraction features over time. DPPC decreased HA hydrophobicity, as increasing DPPC concentrations caused the HA hydrophobic centers to be occupied such that the hydrophobic tails of DPPC are visualized as being directed toward HA chains, whereas hydrophilic head groups toward the solution with the number of hydrophobic interactions between HA atoms decreased. When HA Rg
decreases, the hydrophilic head groups already associated with HA begin to interact with head groups of other already associated DPPC molecules. Short HA chains displaying linear unfolded forms tend to roll up into globular forms which tend to associate with PL, as HA hydrophobic sites are available and consequently unable to organize into large supramolecular PL–HA spatial networks. On the other hand, longer chains adsorb PL, causing Rg
decreases which themselves can prevent contribution to network formation. Only large HA chains exhibit the simultaneous capability to crosslink through intra- and inter-chain hydrophobic centers and to absorb to PL head groups that form surface-active complexes. Such hydrophobic and electrostatic phenomena are notable with DPPC [25
]. An additional consequence of HA globalization secondary to increasing PL concentration is the worsening of synovial fluid viscoelastic properties [42
], further reflecting system dysfunction.
Our data indicated that the average binding energy changed proportionally to the number of hydrogen bonds and HA chains were adsorbed at vesicle surfaces at high
ratios (Figure 15
). This mechanism was even more pronounced, as DPPC created a vesicle encapsulating HA in its interior and excluded those chains from crosslinking opportunities [43
]. At pH = 7.0, no significant surface area changes were observed among DPPC concentrations when compared to other pH values that exhibit a variable surface area. Because this phenomenon is mostly due to DPPC minimizing the hydrophobic contact area with water, the orientation was strongly dependent on concentration; and as more DPPC molecules were present, micellization forces overcame HA and DPPC interactions to minimize the hydrophobic interactions in the system.
The PL increase in osteoarthritic synovial fluid results in a higher number of aggregated forms, such as micelles and reverse micelles, which may not provide a proper response to certain external loads due to the lack of HA networks preventing micellar collapse. Significantly, mechanical damage to the continuity of articular cartilage surfaces can be manifest as an increased synovial fluid pH and PL molecule concentration without significantly affecting Rg. In aqueous solutions such as synovial fluid, non-polar groups of atoms tend to aggregate, since aggregation reduces the number of water molecules that must participate in an ordered, and entropically unfavorable, interfacial cage structure. This water-mediated hydrophobic effect can be interpreted as an attractive hydrophobic interaction between nonpolar groups of atoms. Statistically, hydrophobic interactions are rather long-ranged, extending up to 7 Å, such that a space deficiency for water molecules exists between the interacting groups, leading to an unfavorable vacuum and thus strong attractive forces.
PL composition, including molecular saturation levels and fatty acid chain length, is significantly transformed during osteoarthritic pathogenesis. Although unsaturated, rather than saturated, phosphatidylcholines have been linked to the lubrication of synovial joint organ systems [14
], DPPC should not yet be abandoned as an adjuvant candidate in certain clinical situations. Like sphingomyelin [29
], whose hydrophobic chains are highly saturated, prone to intermolecular hydrogen bonding and lipid raft formation, and exhibit a higher phase transition temperature, DPPC may be important for mitigating heat exchanges associated with the deformation cycles of diseased articular surfaces due to its own higher gel–liquid crystal transition temperature [15
]. Because PLs are present on articular cartilage surfaces in layer and vesicle form, phase transition temperature may be important, in addition to lateral layer heterogeneity, for self-assembly capabilities in association with HA through perturbation forces typically encountered post-treatment [29
]. HA associates with the outer shell of DPPC in vesicle form in bulk solution, which can be deposited on to surfaces—easily deformed into other structures, such as multilayers—or removed altogether under higher loads. Further, it has been recognized that when PL progresses deeper into its solid phase, better lubrication is often created, although some PLs in liquid disordered phases may be preferable simply based upon anticipated loading conditions [49
]. Such post-operative loading after cartilage repair is a complex variable requiring further assessment [51
We remain optimistic that tribological surgical adjuvants can be deployed toward improving chemomechanotransductive environments at articular cartilage repair sites sufficient to warrant wear mitigation studies. In this work, we investigated how the small-sized HA faction behaves with increasing pH values and PL concentration as observed in osteoarthritis. While we have demonstrated that low molecular weight HA, by interacting with PL, would be ineffective for facilitated lubrication and deteriorates synovial fluid viscoelastic properties by nonparticipation in the crosslinking of long-chain HA molecules, DPPC may play a useful adjuvant role in certain clinical situations. An additional evaluation of HA and PL interactions is required to determine in vivo molecular benchmarks that would facilitate optimal absorption–desorption kinetics at surgically repaired surfaces toward re-establishing an effective lubricating layer [52
] to enhance tissue rescue. The framework deployed in this study, namely to evaluate phenotypic changes that occur commensurate with the disease state, can be used to optimize repair site durability and to estimate the point-of-no-return at which articular cartilage lesions should remain relegated to palliative care.