Three-dimensional imaging analysis of CAD/CAM custom-milled versus prefabricated allogeneic block remodelling at 6 months and long-term follow-up of dental implants: A retrospective cohort study.

AIM: This retrospective cohort study aimed to volumetrically investigate the bone stability rate of prefabricated allogeneic bone blocks (PBB) and computer-aided design (CAD)/computer-aided manufacturing (CAM) custom-milled allogeneic bone blocks (CCBB) for ridge augmentation. MATERIALS AND METHODS: Nineteen patients were treated with 20 allografts: 11 CCBB, 9 PBB; 10 in the maxilla and 10 in the mandible. Clinical treatment history and cone beam computed tomography scans before surgery (t0), directly after graft surgery (t1) and after 6 months of healing prior to implant insertion (t2) were evaluated using a three-dimensional evaluation software for absolute bone volume, stability as well as vertical and horizontal bone gain. Furthermore, the inserted implants were analysed for survival, marginal bone loss (MBL) and complications for a mean follow-up period of 43.75 (±33.94) months. RESULTS: A mean absolute volume of 2228.1 mm3 (±1205) was grafted at t1. The bone stability rate was 87.6% (±9.9) for CCBB and 83.0% (±14.5) for PBB. The stability was higher in the maxilla (91.6%) than in the mandible (79.53%). Surgery time of PBB was longer than for CCBB (mean Δ = 52 min). The survival rate of the inserted implants was 100% with a mean MBL of 0.41 mm (±0.37). CONCLUSION: The clinical performance of both allograft block designs was equally satisfactory for vertical and horizontal bone grafting prior to implant placement. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov: NCT06027710.

Marginal changes at bone-level implants supporting dental prostheses with or without intermediate standardised abutments after 36 months: Randomised controlled clinical trial.

OBJECTIVE: The objective of this study is to evaluate the changes at marginal bone level at implants restored with screw-retained prosthesis connected directly to the implants or with an intermediate abutment, after 3-year follow-up. MATERIAL AND METHODS: Thirty-six partially edentulous patients received 72 implants. Each patient received 2 implants and a 2-4-unit screw-retained implant-prosthesis. The test group implants received a screw-retained prosthesis connected directly to the implant shoulder, the control group prosthesis were connected through a 3-mm standardised intermediate abutment. Clinical and radiological data were recorded at baseline and at 6-, 12-, and 36-month follow-up. RESULTS: At 36 months, the mean marginal bone loss was 0.13 ± 0.18 mm for the control group and 0.20 ± 0.24 for the test group, with no significant differences between groups (p > .05). Clinical variables (Probing Pocket Depth, Bleeding on Probing and Plaque Index) at 36 months also showed no significant difference between groups. Minor complications frequency was 6.7% in the control group and 5.3% in test group. None of the groups suffered from mayor complications. Patient Reported Outcomes (PROs) showed a General Satisfaction mean score in the control group of 9.40 (SD 0.82) and 9.37 (SD 1.06) in the test group with no significant differences between groups. CONCLUSIONS: Bone-level implants restored with screw-retained partial prostheses with or without intermediate abutments showed similar radiographic and clinical outcomes after 3 years.

Marginal changes at bone-level implants supporting fixed screw-retained partial implant prostheses with or without intermediate standardised abutments: 1-year results of a randomised controlled clinical trial.

OBJECTIVE: To compare marginal changes at bone-level implants restored with screw-retained implant prosthesis with or without intermediate standardised abutments, after 1 year of follow-up. MATERIALS AND METHODS: Thirty-six partially edentulous patients received 72 implants. Each patient received 2 implants and a 2- to 4-unit screw-retained implant-prosthesis. The test group received implants consisting of a screw-retained prosthesis connected directly to the implant shoulder, while the prostheses in the control group were connected through a 3-mm standardised intermediate abutment. Clinical and radiological data were recorded at baseline and at 3, 6 and 12 months in follow-up visits. RESULTS: At 12 months, the marginal bone loss was 0.17 ± 0.24 mm for the test group (19 patients) and 0.09 ± 0.15 mm for the control group (17 patients), with no statistically significant differences (p > .05). The mean probing pocket depth was 2.96 mm ± 0.46 for the test group and 2.86 ± 0.62 mm for the control group. The test and control groups showed bleeding on probing levels of 18.86 ± 14.12% and 13.73 ± 17.66%, respectively. All patients scored below 25% on the plaque index levels. CONCLUSIONS: Restoration of bone-level implants with fixed screw-retained partial prostheses with or without intermediate abutments presented similar radiographic and clinical outcomes after 1 year.

Periodontitis as a Preterm Birth Risk Factor in Caucasian Women: A Cohort Study.

PURPOSE: The aim of this study was to evaluate the association between periodontitis and preterm birth in a Spanish Caucasian population, based on clinical and biochemical outcomes. Epidemiological studies have suggested that periodontitis is a potential risk factor for preterm birth. However, other studies have shown high heterogeneity in their results. Some factors such as number of evaluations during pregnancy, sample size, study population and maternal age may have an impact on the variability of the result. METHODS AND MATERIALS: This cohort study enrolled 158 pregnant women, 39 with periodontitis and 119 without periodontitis. All pregnant women were evaluated in the first, second and third trimester. RESULTS: Statistically significant differences were found in periodontal parameters between both groups, but no statistically significant differences were found in biochemical parameters during pregnancy. The duration of pregnancy in healthy patients was 38.78 ± 4.49 weeks, and in patients with periodontitis 37.81 ± 4.89 weeks, with no statistical difference (p > 0.05). This showed that periodontitis was not associated with preterm birth in a Spanish Caucasian cohort. CONCLUSION: In this study, periodontitis stage II, grade B, was not statistically associated with preterm birth. Pregnancy is a short period of time in order to evaluate long-term oral systemic infections. Adverse pregnancy outcomes are more difficult to occur. Thus, since pregnancy timing average cannot be changed, the stages of periodontal disease (initial, moderate, advanced) could be another factor to study.

Influence of abutment height and implant depth position on interproximal peri-implant bone in sites with thin mucosa: A 1-year randomized clinical trial.

OBJECTIVES: The aim of this RCT was to assess radiographically the effect of abutment height and depth of placement of platform-switched implants on interproximal peri-implant bone loss (IPBL) in patients with thin peri-implant mucosa. MATERIAL AND METHODS: Thirty-three patients received one prosthesis supported by two implants replacing at least two adjacent missing teeth (66 implants). Patients were randomly allocated and implant insertion depth adapted to abutment height groups (3 mm height group the implants were placed 2 mm subcrestally; 1 mm height group, equicrestally). Clinical and radiological measurements were performed at 3, 6 and 12 months after surgery. Interproximal bone-level changes were compared between treatment groups using repeated measures mixed ANOVA. The association between IPBL and categorical variables was also analyzed. RESULTS: The mean IPBL in 1 mm abutment group was 0.76 ± 0.79 mm at 3 months, 0.92 ± 0.88 mm at 6 months, and 0.95 ± 0.88 mm at 12 months, while in the 3 mm abutment group was 0.06 ± 0.21, 0.07 ± 0.22 mm, and 0.12 ± 0.33 mm, respectively. Significant differences between both groups were observed at every time point. When the influence of patient characteristics and clinical variables was analyzed, no statistically significant differences were also observed. CONCLUSIONS: The use of long abutments, in combination with subcrestal implant position in sites with thin mucosa, led to lower IPBL in comparison with the use of short abutments.

Effect of abutment height on interproximal implant bone level in the early healing: A randomized clinical trial.

OBJECTIVE: The aim of this randomized clinical trial was to compare the effect on the interproximal implant bone loss (IBL) of two different heights (1 and 3 mm) of definitive abutments placed at bone level implants with a platform switched design. MATERIAL AND METHODS: Twenty-two patients received forty-four implants (6.5-10 mm length and 3.5-4 mm diameter) to replace at least two adjacent missing teeth, one bridge set to each patient-two implants per bridge. Patients were randomly allocated, and two different abutment heights, 1 and 3 mm using only one abutment height per bridge, were used. Clinical and radiological measurements were performed at 3 and 6 months after surgery. Interproximal bone level changes were compared between treatment groups. The association between IBL and categorical variables (history of periodontitis, smoking, implant location, implant diameter, implant length, insertion torque, width of keratinized mucosa, bone density, gingival biotype and antagonist) was also performed. RESULTS: At 3 months, implants with a 1-mm abutment had significantly greater IBL (0.83 ± 0.19 mm) compared to implants with a 3-mm abutment (0.14 ± 0.08 mm). At 6 months, a greater IBL was observed at implants with 1-mm abutments compared to implants with 3-mm abutments (0.91 ± 0.19 vs. 0.11 ± 0.09 mm). The analysis of the relation between patient characteristics and clinical variables with IBL revealed no significant differences at any moment except for smoking. CONCLUSIONS: Abutment height is an important factor to maintain interproximal implant bone level in early healing. Short abutments led to a greater interproximal bone loss in comparison with long abutments after 6 months. Other variables except smoking showed no relation with interproximal bone loss in early healing.

Influence of Abutment Height on Maintenance of Peri-implant Crestal Bone at Bone-Level Implants: A 3-Year Follow-up Study.

In this 3-year follow-up study, peri-implant bone loss at bone-level implants was evaluated with two definitive abutment heights: 1 mm and 2.5 mm. Peri-implant bone loss was defined as the distance between the implant shoulder and the first bone-to-implant contact from the time of loading to the 36-month follow-up, estimated using periapical radiographs. The bone loss was increased at the time of follow-up, to 1.30 mm (95% confidence interval [CI]: 0.70-1.89 mm; SD = 1.89) and 0.33 mm (95% CI: 0.11-0.55; SD = 0.59) at 36 months in short and long abutments, respectively. Placement of short abutments induced higher peri-implant bone loss at bone-level implants during a peri-implant recall program.

Peri-implant soft tissue analyses comparing Ti and ZrO2 abutments: an animal study on beagle dogs.

OBJECTIVE: The aim of this study was to evaluate the soft tissue histomorphometric composition around implant abutments comparing two different materials, titanium (Ti) and zirconia (ZrO2 ). MATERIAL AND METHODS: Twelve implants were placed at bone level in the mandible of six beagle dogs (one in each side). At the same day of surgery one titanium abutment was screwed to the implant in one side (control group) and a zirconia abutment was screwed in the contralateral side. Nine months after implant/abutments placement, animals were sacrificed for histological analysis. Descriptive analysis was calculated for each variable and Wilcoxon test was applied to evaluate histomorphometric variables. RESULTS: At the end of the study the soft tissue dimension at Ti and ZrO2 were similar in all counterparts: biological width, the length of the barrier epithelium, length of the connective tissue, and the percentage of collagen fibers density. However, the percentage of blood vessels was higher for the Ti in comparison to ZrO2 (5.11% ± 1.70 and 2.23% ± 0.98, respectively [P = 0.016]). CONCLUSIONS: Peri-implant soft tissue histomorphology composition was similar in implant abutments made of ZrO2 and titanium after 9 months of healing.

Enzyme-based microfluidic chip coupled to graphene electrodes for the detection of D-amino acid enantiomer-biomarkers.

An electrochemical microfluidic strategy for the separation and enantiomeric detection of D-methionine (D-Met) and D-leucine (D-Leu) is presented. These D-amino acids (D-AAs) act as biomarkers involved in relevant diseases caused by Vibrio cholerae. On a single layout microfluidic chip (MC), highly compatible with extremely low biological sample consumption, the strategy allowed the controlled microfluidic D-AA separation and the specific reaction between D-amino acid oxidase (DAAO) and each D-AA biomarker avoiding the use of additives (i.e., cyclodextrins) for enantiomeric separation as well as any covalent immobilization of the enzyme into the wall channels or on the electrode surface such as in the biosensor-based approaches. Hybrid polymer/graphene-based electrodes were end-channel coupled to the microfluidic system to improve the analytical performance. D-Met and D-Leu were successfully detected becoming this proof-of-the-concept a promising principle for the development of point-of-care (POC) devices for in situ screening of V. cholerae related diseases.

Graphene oxide nanoribbon-based sensors for the simultaneous bio-electrochemical enantiomeric resolution and analysis of amino acid biomarkers.

In this work, a straightforward in-situ measurement of L and D-amino acids (AAs) has been developed using disposable graphene oxide nanoribbon (GON) screen printed electrodes. For that, we took advantage of the electroactivity of certain clinically relevant AAs, such as tyrosine (Tyr) and methionine (Met), which are involved in important bacterial diseases (Bacillus subtilis and Vibrio cholera, respectively). The strategy is based on a dual electrochemical and enzymatic approach. The D-AA with the class enzyme D amino acid oxidase (DAAO) generates H2O2. This H2O2 is simultaneously detected with the L-AA, electroactive molecule by differential pulse voltammetry (DPV). These GON disposable platforms use just 50 µL of sample and a total analysis time of 360 s. Both L and D enantiomers calibration and quantitative analysis were explored and were simultaneously detected with accuracy and precision in urine samples. Any interference of uric acid and other electroactive AAs was noticed. This proposed electrochemical GON-based enantiomeric bio-sensor becomes a highly promising tool as future point of care for fast and reliable early diagnosis of diseases related to the presence of D-AAs.

Tips for the functionalization of nanoparticles with antibodies.

Multiple antibody immobilization methodologies have been developed for several applications including affinity chromatography, immunosensing, and drug delivery. Most of them have been carried out without considering the orientation of the antigen binding site of the antibody, or after the chemical modification of the antibody. An efficient immobilization to improve the biological activity of the antibody is one of the key fundamental issues to pursue. A simple and effective methodology for well-oriented covalently immobilization of antibodies on nanoparticles is reported in this chapter.

Oriented covalent immobilization of antibodies onto heterofunctional agarose supports: a highly efficient immuno-affinity chromatography platform.

The development of new bioconjugates formed by one antibody optimally bound (through its Fc region) to fairly inert solid surfaces is of primary relevance in immuno-affinity chromatography. Immunoglobulins G (IgG) have a Fc region very rich in histidine (His) residues. In this way, immobilization of IgGs on heterofunctional metal chelate-glyoxyl supports (Ag-Me(2+)/G) takes place in two steps: firstly the antibodies are conjugated to the support via His-metal coordination bonds. Secondly, their incubation under alkaline condition promotes an intramolecular covalent attachment between lysine residues at the Fc region and glyoxyl groups on the support surface. The IgG that recognizes as antigen the HRP (antiHRP-IgG) has been conjugated to Ag-Me(2+)/G supports. The resulting bioconjugate is highly inert and able to specifically bind the antigen (HRP) without significant unspecific binding of any other proteins, resulting in an excellent HRP purification platform. The binding activity of this bioconjugate has been optimized by controlling the antibody distribution throughout the bead's surface in order to avoid high antibody densities that led to a low binding activity of the antibodies. The optimal antibody distribution has been achieved when these proteins were slowly immobilized on Ag-Cu(2+)/G in presence of imidazole. This bioconjugate was able to bind up to 1.5 moles of antigen per mole of antibody, only 1.3-fold less than the antibody in solution. Hence, we have been able to develop an optimal protocol to prepare bioconjugated composites in an oriented and irreversible fashion which results in highly efficient and specific surfaces for the exclusive biological recognition.

Taking advantage of unspecific interactions to produce highly active magnetic nanoparticle-antibody conjugates.

Several strategies for linking antibodies (Abs) through their Fc region in an oriented manner have been proposed at the present time. By using these strategies, the Fab region of the Ab is available for antigen molecular recognition, leading to a more efficient interaction. Most of these strategies are complex processes optimized mainly for the functionalization of surfaces or microbeads. These methodologies imply though the Ab modification through several steps of purification or the use of expensive immobilized proteins. Besides, the functionalization of magnetic nanoparticles (MNPs) turned out to be much more complex than expected due to the lack of stability of most MNPs at high ionic strength and non-neutral pH values. Therefore, there is still missing an efficient, easy and universal methodology for the immobilization of nonmodified Abs onto MNPs without involving their Fab regions during the immobilization process. Herein, we propose the functionalization of MNPs via a two-steps strategy that takes advantage of the ionic reversible interactions between the Ab and the MNP. These interactions make possible the orientation of the Ab on the MNP surface before being attached in an irreversible way via covalent bonds. Three Abs (Immunoglobulin G class) with very different isoelectric points (against peroxidase, carcinoembryonic antigen, and human chorionic gonadotropin hormone) were used to prove the general applicability of the strategy here proposed and its utility for the development of more bioactive NPs.

The epitope for the polyol-responsive monoclonal antibody 8RB13 is in the flap-domain of the beta-subunit of bacterial RNA polymerase and can be used as an epitope tag for immunoaffinity chromatography.

Polyol-responsive monoclonal antibodies (PR-mAbs) are useful for the purification of proteins in an easy, one step immunoaffinity step. These antibodies allow for gentle purification of proteins and protein complexes using a combination of a low molecular weight polyhydroxylated compound (polyol) and a nonchaotrophic salt in the eluting buffer. mAb 8RB13 has been characterized as one of these PR-mAbs and has been used to purify RNA polymerase from five species of bacteria. Here the epitope for 8RB13 has been identified as PEEKLLRAIFGEKAS, a sequence that is highly conserved in the ß-subunit of bacterial RNA polymerase. This sequence is located in the "beta-flap" domain of RNA polymerase (and essentially comprises the "flap-tip helix"), an important binding site for sigma70. This location explains why only the core RNAP is purified using this mAb. This amino acid sequence has been developed into an epitope tag that can be used to purify a target protein from either bacterial or eukaryotic cells when genetically fused to a protein of interest.

Selective adsorption of small proteins on large-pore anion exchangers coated with medium size proteins.

A new anion exchanger support has been designed for the selective adsorption of small proteins. This has been achieved activating an aminated support with glutaraldehyde and further coating the support surface with bovine serum albumin (BSA). In this support, "wells" are generated by two neighborhoods BSA molecules, on the bottom of those "wells" glutaraldehyde groups are exposed out ready to react with small molecules that have a size small enough to be accommodated between two BSA molecules on the pre-existing support. However, the BSA surface was not inert enough adsorbing many proteins, thereby reducing the selectivity of the system. A further solution was coating the immobilized BSA molecules with dextran, reducing the adsorption of protein on the BSA surface. This new matrix has been evaluated in the selective adsorption of the very small beta-lactoglobulins and alpha-lactalbumin from dairy whey, achieving the selective adsorption of both small proteins while other larger proteins from dairy whey remained in the supernatant. Moreover, a protein crude extract has been offered to the new matrix, and only small proteins could be adsorbed on the support (as probed by gel filtration). Thus this amino-glutaraldehyde-BSA-dextran-Sepharose is a matrix that may be used to selectively ionically adsorb proteins that were smaller than BSA (62 kDa). This strategy may be used for any other kind of adsorbing groups (chelating agents, boronic acid, etc.), or using proteins with different sizes to coat the support, designing tailor-made supports that may permit the fractioning of proteins following their sizes and by adsorption/desorption on different matrices.

Covalent immobilization of antibodies on finally inert support surfaces through their surface regions having the highest densities in carboxyl groups.

The correct immobilization of antibodies is one of the most critical steps in the preparation of immunosensors and immunochromatography matrices. In addition, the final support has to be chemical and physically inert to avoid the unspecific adsorption of proteins that can reduce the sensitivity of the biosensor or the purification achieved by the chromatography. The solution to both problems is one of the major challenges in the field. Here, we have presented two different novel and simple alternatives to have the unmodified antibody anionically exchanged to a support, further covalently immobilized with more than 90% of the antibodies bonded to the support by the four subunits, retaining a high functionality and giving a final "inert" surface. The first solution was the use of supports having a low superficial density of amino groups activated with glutaraldehyde. Here, the inertness was achieved by the use of a very low density of amino groups, unable to adsorb proteins at 100 mM sodium phosphate, while immobilization proceeds mainly via a first adsorption of the antibody and a further reaction with the glutaraldehyde groups. The second solution implies the design of a novel support (amino-epoxy). This support again produces a first ionic exchange of the antibody on the support and a further reaction with the epoxy groups, but because the epoxy groups can be finally blocked with aspartic groups (annulling the charge), the initial density of amino-epoxy groups can be as high as possible. Both systems permitted the correct and oriented immobilization of IgG. The immobilized antibody showed high-functionality (65-75%) and a final inert support surface. This immobilized antibody (antiperoxidase) was able to capture fully specifically HRP contaminating a protein crude extract from E. coli.

Oriented covalent immobilization of antibodies on physically inert and hydrophilic support surfaces through their glycosidic chains.

Immobilization of antibodies by their oxidized sugar chain on aminated supports is a very efficient methodology to have a properly oriented antibody. However, these supports may behave as anionic exchangers, producing the unspecific adsorption of other proteins and reducing the selectivity of the system. To overcome this problem, we have proposed two solutions based in tailor-made support surfaces to immobilize antihorseradish peroxidase (HRP). The first solution was the use of supports having a very low amount of amino groups. These amino groups need to be very reactive with the aldehyde groups generated in the protein sugar chains to be efficient. Using supports having 7 micromol EDA/g (e.g., ethylenediamine modified glyoxyl-agarose), the antibody may be immobilized, keeping over 90% of the anti-HRP functionality. Second, by mixing amino groups and carboxylic groups, a neutral surface of the support has been generated. Again, this support has been unable to adsorb proteins while oxidized anti-HRP could be immobilized, giving functional anti-HRP antibodies. Both preparations retained 100% functionality after 2 months of storage at 4 degrees C. This way, the tailoring of the support surfaces has permitted solving some limitations of the immobilization of sugar-chain oxidized antibodies on primary amino supports.

Mixed ion exchange supports as useful ion exchangers for protein purification: purification of penicillin G acylase from Escherichia coli.

A support having similar amounts of carboxymethyl and amino groups has been prepared and evaluated as an ion exchanger. It has been found that this support was able to adsorb a high amount of protein from a crude extract of proteins (approximately 55%) at pH 5. Moreover, it was able to adsorb approximately 60% of the protein that did not become adsorbed on supports bearing just one kind of ionic groups. The use of divalent cations reinforced the adsorption of proteins on these supports. These results suggest that the adsorption of proteins on supports bearing almost neutral charge is not driven by the existence of opposite charges between the adsorbent and the biomacromolecule but just by the possibility of forming a high number of enzyme-support ionic bonds. This support has been used to purify the enzyme penicillin G acylase (PGA) from Escherichia coli. PGA was not significantly adsorbed at any pH value on either amino- or carboxyl-activated supports, while it can be fully adsorbed at pH 5 on this new carboxyl-amino matrix. Thus, we have been able to almost fully purify PGA from crude extracts with a very high yield by using these new supports.

Solid phase proteomics: dramatic reinforcement of very weak protein-protein interactions.

Very weak protein-protein interactions may play a critical role in cell physiology but they are not easily detectable in "in vitro" experiments. To detect these weak interactions, we have developed a strategy that included: (a) design of a rapid and very effective crosslinking of protein-protein complexes with poly-functional reagents; (b) selective adsorption of very large proteins on lowly activated ionic exchangers, based on the need of a multipoint physical adsorption to incorporate the proteins into the matrix; (c) purification by selective adsorption of protein-protein complexes formed by strong protein-protein interactions, via selective adsorption of the complexes on lowly activated ionic exchangers via multi-protein physical adsorption and leaving the non-associated proteins in the solution; (d) reinforcement of very weak protein-protein interactions by selective adsorption of the complex on lowly activated ionic exchange supports via a synergetic cooperation of the weak protein-protein interaction plus the interactions of both proteins with the support enabling the almost full shifting of the equilibrium towards the association position; (e) control of the aggregation state of proteins like BSA, formed by weak protein-protein interactions. In this last case, it seems that the interaction of the protein molecules placed on the borders of the aggregate with the groups on the support partially stabilizes the whole aggregate, although, some molecules of the aggregate cannot interact with the support. The size of the aggregates may be defined by controlling the concentration of ionised groups on the support: the less activated the supports are, the bigger the complexes. In this way, solid-phase proteomics could be a very interesting tool to detect weak protein-protein interactions.

Adsorption behavior of bovine serum albumin on lowly activated anionic exchangers suggests a new strategy for solid-phase proteomics.

Diluted solutions of bovine serum albumin (BSA) (e.g., 0.1 mg /mL) do not form detectable protein large aggregates. Using gel-filtration experiments, we determined that a diluted solution of BSA is 97% monomeric BSA and 3% dimeric. The adsorption of this diluted BSA on highly activated anionic exchangers (e,g., having 40 micromol/wet g) keeps this mainly monomeric form. When supports activated with 2 micromol/wet g are used, only dimers become adsorbed to the support, accounting for 100% of the offered BSA. When the diluted BSA solution is offered to very mildly activated anionic exchangers (even only 0.125 micromol/wet g), an unexpected adsorption of most of the BSA on the support was also observed. These very slightly activated supports are only able to adsorb very large proteins or very large protein-protein complexes, larger than BSA dimers. In fact, a rapid cross-linking of the adsorbed BSA with dextran-aldehyde reveals the formation of very large BSA-BSA complexes with molecular mass higher than 500 000 Da, complexes that may be observed for soluble BSA with very high concentrations but are not detectable at 0.1 mg/mL. Moreover, the size of the aggregates strongly depends on the concentration of the ionized groups on the support: the less activated the supports are, the higher the sizes of the complexes. It seems that the interaction of the BSA molecules on the margins of the BSA aggregate with the groups on the support may stabilize the whole protein aggregate, although some components are not interacting with the support. Aggregates could account for more than 40% of the BSA in the solution after 50 h of incubation. However, only these large BSA aggregates were adsorbed in the support.