Histological evaluation of periprosthetic implant associated tissue, retrieved from patients undergoing hip replacement surgery due to PE induced osteolysis, indicated chronic inflammatory infiltrates (). Immunophenotyping demonstrated that these infiltrates consisted predominantly of CD68+ macrophages (MΦ), with few T and B cell infiltrates (). Histiocytes and foreign body giant cells engulfing birefringent PE particles, generated by wear and tear of the polymeric implant, were also observed ().
Periprosthetic inflammatory reaction to micron and sub-micron size modified n-alkane polymers.
Ultrastructural analysis revealed the presence of micron and sub-micron PE particles in the periprosthetic inflamed tissue (), similar to the one derived from chemically synthesized carbonyl modified alkane polymers (). Micron size PE particles were mostly extracellular and observed among collagen fibers (). Sub-micron PE particles on the other hand were phagocytosed by local antigen presenting cells ().
The presence of PE material at the site of inflammation raised the question of the molecular basis for alkane polymers recognition by the immune system. As initial analysis, a biophysical evaluation by fourier transformed infra-red spectroscopic (FTIR) 
was performed on pre and post-implant material to determine possible alkane polymer modification that could explain the loss of its bio-inert properties (). Analysis was performed on (i) pre-implant PE, corresponding to the actual biological implant before surgery, and (ii) post-implant PE, which was retrieved from the site of osteolysis at the time of revision surgery (). Post-implant material was purified according to scheme reported in . Lipid, sterols and protein analysis determined that the preparation was void of detectable contaminants as previously reported 
(FTMS analysis and assignment of molecular formulas further confirm purity from lipids and protein contaminants ().
Modified alkane polymers purified from periprosthetic material increase their number of carbonyl groups.
Mass spectroscopy analysis of n-alkane polymers purified from post implant PE indicates extensive oxidation.
FTIR signature peaks at 2900–2800 cm−1 wave numbers showed the asymmetric stretch of the alkane backbone and a skeletal vibration of the same backbone was registered at 715 cm−1 for all the four samples. The alkane backbone also exhibited a predicted deformation at around 1470 m−1 in each case. Differences in spectra were noted between the pre and post implant PE material. A dramatic increase in the amount of carbonyl (carboxylic, ketonic, aldehydic and ester), amide and alcohol groups was observed in PE polymers prepared from post-implant material (). The increase in the amount of carbonyl groups (236% ) is due to an “in situ” oxidative process. This oxidative process may be mediated by enzymes released from activated DC, MΦ and osteoclasts, or may occur in the endosomal compartment of local antigen presenting cells engorged with PE polymers ().
To further define the molecular composition of the alkane polymers retrieved from the post-implant material a Fourier transform ion cyclotron mass spectroscopy FT/MS in MALDI (matrix assisted laser desorption ionization) mode was performed. FT/MS is so far the most sophisticated technique to identify molecular composition since it combines the most advanced Ion Trap and Fourier Transform Ion Cyclotron Resonance technologies into a single instrument with unprecedented analytical power. Ultra-high resolution and sensitivity coupled with sub mass prediction accuracy allow determining elemental composition. Post-implant material was obtained as described in and short chain alkane polymers were further purified from the nano and micron size PE particles by centrifugation through a 10,000 membrane cut-off. Pre-implant material was similarly prepared. FTMS analysis indicated the presence of several small sized PE polymers in the post-implant material with a molecular mass of 400 to 2000 (). Differently from the pre-implant PE, where a clear polymer-associated envelope was observed, in the post implant material this signature was lost due to extensive oxidation. Chemical formulas were assigned for each molecular species observed in the post-implant PE. Assignment of each pick was consistent with an alkane structure back bone bearing several side chain modifications (). None of the assigned formulas could account for the presence of proteins, peptides, lipoproteins or lipopeptides further confirming the purity of the post-implant preparation. Most of the high molecular weight alkane polymers presented extensive addition of carboxyl groups ( and Figure S1
). In the lower molecular weight mass several polymers could be observed with a low degree of oxidation similar to the MS/MS profile, of mPE polymers (chemically synthesized PE polymers with few side chain modifications (hydroxyl and carboxyl functional groups) ( and Figure S2
). In both samples two major clusters of peaks in the 600 and 800 mass range could be observed (). Further analysis by TOF-TOF MALDI 
fragmentation indicated an identical fragmentation pattern for both mPE and post implant PE. In both cases the 617 peak fragmented into 428 and 572 m/z, while the 806 peak fragmented into 617 (Figure S3
). We therefore concluded that polymers of different sizes and amount of oxidation were present at the site of inflammation associated with a strong innate immune response.
The observation of modified PE material at the site of inflammation associated with cellular infiltrates prompted us to evaluate whether exposure to the modified polymers would activate antigen presenting cells. To test for this, we used pre-implant PE, post-implant PE, and chemically synthesized PE polymers without (unPE) or with side chain modification (hydroxyl and carboxyl functional groups) (mPE) matching several of the modification observed in the post-implant PE (). Dendritic cells (DC) cultured for 48 hours in the presence of 50 μg/ml of each compound were evaluated for surface MHC class II expression. An up-regulation of MHC class II molecules was observed in post-implant and mPE treated cells only, to level similar to what observed with the lipopeptide Pam2CSK4 (). DC activation was further confirmed by a significant increase in IL-12 secretion determined by ELISA (). We concluded that the hydroxyl and carboxyl modified mPE polymers and post-implant PE can interact with DC and initiate an inflammatory response.
Modified alkane polymers induce DC activation and IL-12 secretion.
A strong and rapid initiation and activation of the innate immune response is generally achieved through engagement of members of the toll like receptor (TLR) family 
. Hence, we investigated further whether any of the TLRs were actually involved in the recognition of the mPE as well as post-implant modified alkane polymers. Human 3T3 HEK cell lines stably expressing TLR1/2, TLR2, TLR3 and TLR4 genes respectively were transfected with a plasmid encoding the luciferase reporter gene under the control of the NF-κΒ enhancer element. Thus, NF-κΒ activation by pro-inflammatory stimuli would, in turn, up-regulate luciferase production. Each TLR cell line was assayed for activation with different PE preparations, chemically synthesized (unPE and mPE) or alkane polymers prepared from the pre implant and post implant material 
. Each assay also included a well known specific positive ligand for each respective TLR (). Both mPE and post-implant preparations induced luciferase up-regulation in both TLR1/2 and TLR2 transfectants, but not in the TLR3 and TLR4 cell lines (). Thus, we conclude that the carbonyl modified PE polymers cause a strong activation of the innate immune system and initiate an inflammatory cascade through the TLR1/2 receptor activation pathway.
Modified alkane polymers induce activation of TLR-1 and TLR-2 signaling pathways.
To further analyze the TLR2 binding activity of modified alkane polymers a soluble form of human recombinant TLR2 (extracellular domain Glu 21-Leu 590) was utilized in a binding assay that monitored changes in fluorescence intensity (λexcitation
277 nm and λemission
335 nm) of a tyrosine present in the TLR2 binding grove (Tyr 326) 
. Changes in the binding grove environment, due to ligand occupancy would change the Tyr fluorescence intensity and the wavelength of its fluorescence emission. Soluble TLR-2 was incubated with increasing concentrations of each polyethylene derivatives; mPE (as a mixture of hydroxyl and ester bond alkane polymers), PE (unmodified alkane polymers) pre and post implant PE and the positive control, Pam2CSK4 lipopeptide (known to be a specific ligand for TLR-2) 
. The emission scans (between 290 and 420 nm) were collected for each complex separately and the change in maximum fluorescence signal at 335 nm (due to tyrosinate ion) was used to generate the binding curves, after subtracting the contribution of the free protein (in the absence of any compounds). The normalized fluorescence data were fitted to a hyperbolic function using the software GraphPad Prism 4 ().
Direct binding of oxidized alkane polymers to soluble TLR-2 molecules.
The steady-state intrinsic fluorescence emission of TLR-2 strongly increased to ligand saturability upon the addition of mPE. Since mPE comprise a random mixture of hydroxyl and carboxyl alkanes (Figure S2
) a binding Kd was calculated for both components. For hydroxyl and carboxyl modified alkanes the binding Kd was 34.18 μM and 35.75 μM respectively (). Thus, the predicted Kd range for the modified alkane mixture was very close to the Pam2CSK4 positive control (). Importantly, the alkane binding to soluble TLR-2 was inhibited following incubation with the monoclonal antibody anti human TLR2 (clone 383936 R&D Systems) which is known to prevent ligand access to the TLR2 binding groove 
. On the other hand the affinity of binding for non-modified alkane polymers (unPE) and pre-implant material was much lower than the one reported for the oxidized ones (). Post-implant PE, a wide mixture of non-oxidized and oxidized alkanes also showed a saturable binding even though the blend of different alkane species did not allow for Kd mesurement to be calculated. Nevertheless, the binding was completely blocked by addition of the TLR2 mAb (). In general a positive correlation was observed between TLR2 binding affinity and amount of oxidation of the alkane polymeric structures () in agreement with the luciferase assay data ().
The crystal structure of TLR1/2 combined with an active bacterial ligand (Pam3
) indicates that the CH2 backbone of the lipid ligand occupies the three hydrophobic active pockets 
. Since FTIR analysis indicated that in the post-implant material alkane oxidation was the prevalent modification of the alkane groups we further characterize the interaction between the oxidized alkane polymers and TLR1/2 receptors. The structural fitness of; (i
) a 24 repetitive units of an alcohol modified alkane polymer (1390 m/z), (ii
) its oxidized form (1406 m/z), (iii
) and an isomer from a post-implant polymer (1333 m/z) into the binding site of TLR-1/2 were evaluated. Molecular docking was performed using as template the 2z7x.pdb structure of the TLR-1/2 in complex with the tri-acylated lipopeptide Pam3CSk4 
(). The mPE and post-implant polymers were designed to have an m/z ratio extrapolated from the FT/MS data ( and Figure S2
) and all the polymers had at least 16 carbons in their backbone based on published structural requirements for a TLR1/2 ligand 
. The polymer structures were merged into the X-Ray structure 2z7x.pdb, such that each individual polymer was superimposed over the original ligand Pam3
. The Pam ligand was then deleted to generate the TLR1/2-mPE complex (). The relative free energy of interaction between each of the new ligands and the TLR1/2 heterodimer was assessed using the MM94FF force field built-in the software SCULPT. Both the hydroxyl-modified alkane (mPE-1390) and the carboxyl form (mPE-1406) as well as the post implant 1333 were predicted to be very good fit based on the relative free energy of interaction as compared to the Pam3
-lipopeptide control (). Thus, additional structural analysis of the interaction between mPE-1390 and TLR1/2 was performed by extracting the amino acids from the binding pocket of the receptor which made contacts with the ligand (). Altogether, the predicted relative free energy of binding, as determined by Vander waals and electrostatic interactions (kcal/mol), together with the analysis of amino acids contacts between the ligand and the receptor clearly indicated a strong probability for hydroxyl-modified alkane polymers to fit appropriately into the TLR1/2 binding groove.
Mass spectroscopy analysis of n-alkane polymers predicted conformations fitting within the hydrophobic pocket of TLR-1/TLR-2 receptor.