Cells resident to precision templated 40-µm pore scaffolds generate small extracellular vesicles that affect CD4+ T cell phenotypes through regulatory TLR4 signaling.

Precision porous templated scaffolds (PTS) are a hydrogel construct of uniformly sized interconnected spherical pores that induce a pro-healing response (reducing the foreign body reaction, FBR) exclusively when the pores are 30-40µm in diameter. Our previous work demonstrated the necessity of Tregs in the maintenance of PTS pore size specific differences in CD4+ T cell phenotype. Work here characterizes the role of Tregs in the responses to implanted 40µm and 100µm PTS using WT and FoxP3+ cell (Treg) depleted mice. Proteomic analyses indicate that integrin signaling, monocytes/macrophages, cytoskeletal remodeling, inflammatory cues, and vesicule endocytosis may participate in Treg activation and the CD4+ T cell equilibrium modulated by PTS resident cell-derived small extracellular vesicles (sEVs). The role of MyD88-dependent and MyD88-independent TLR4 activation in PTS cell-derived sEV-to-T cell signaling is quantified by treating WT, TLR4ko, and MyD88ko splenic T cells with PTS cell-derived sEVs. STAT3 and mTOR are identified as mechanisms for further study for pore-size dependent PTS cell-derived sEV-to-T cell signaling. STATEMENT OF SIGNIFICANCE: Unique cell populations colonizing only within 40µm pore size PTS generate sEVs that resolve inflammation by modifying CD4+ T cell phenotypes through TLR4 signaling.

Adhesion of Staphylococcus epidermidis to biomaterials is inhibited by fibronectin and albumin.

Decades of contradictory results have obscured the exact role of adsorbed fibronectin in the adhesion of the bacterium, Staphylococcus epidermidis, to biomaterials. Here, the ability of adsorbed fibronectin (FN) or bovine serum albumin (BSA) to modulate S. epidermidis adhesion to various biomaterials is reported. FN or BSA was adsorbed in increasing surface densities up to saturated monolayer coverage onto various common biomaterials, including poly(ethylene terephthalate), fluorinated ethylene propylene, poly(ether urethane), silicone, and borosilicate glass. Despite the wide range of surface characteristics represented, adsorption isotherms varied only subtly between materials for the two proteins considered. S. epidermidis adhesion to the various protein-coated biomaterials was quantified in a static-fluid batch adhesion assay. Although slight differences in overall adherent cell numbers were observed between the various protein-coated substrata, all materials exhibited significant dose-dependent decreases in S. epidermidis adhesion with increasing adsorption of either protein (FN, BSA) to all surfaces. Results here indicate that S. epidermidis adhesion to FN-coated surfaces is not a specific adhesion (i.e., receptor: ligand) mediated process, as no significant difference in adhesion was found between FN- and BSA-coated materials. Rather, results indicate that increasing surface density of either FN or BSA actually inhibited S. epidermidis adhesion to all biomaterials examined.

Effects of controlled fibronectin surface orientation on subsequent Staphylococcus epidermidis adhesion.

Several bacterial species, including Staphylococcus aureus and Staphylococcus epidermidis (SE) are known to express cell receptors that bind specifically to surface immobilized or extracellular matrix ligands, such as the protein fibronectin (FN). Yet, few existing studies have examined the effect of protein surface orientation on bacterial adhesion. We report here a substratum modification protocol that allows for the specific orientation of FN molecules on a surface at known levels of surface coverage. Monoclonal antibodies (Mabs), specific to either the COOH-terminus or NH3-terminus of FN, are conjugated to biotin, then immobilized to streptavidin-coated glass substrata. Specific orientation of the bound FN molecules is verified using the same Mabs in an ELISA. Bacterial adhesion of Staphylococcus epidermidis (SE) to FN bound by either its C-terminus or its NH3-terminus was quantified in batch static adhesion assays. Results indicate an increase in SE adhesion to FN-coated surfaces when the FN is bound by its C-terminus (NH3-terminus free), indicating SE receptor-specific adhesion to the FN NH3-terminus. These studies demonstrate that antifibronectin monoclonal antibodies can be used to specifically bind and orient fibronectin on a surface. In addition, adhesion of SE to these model substrata can be controlled by the orientation of the protein.

Inhibition of biofilm formation by monoclonal antibodies against Staphylococcus epidermidis RP62A accumulation-associated protein.

Staphylococcus epidermidis expresses a 140-kDa cell wall-bound protein accumulation-associated protein (AAP) to adhere to and accumulate as a biofilm on a surface. Potentially blocking AAP with a monoclonal antibody (MAb) could reduce or eliminate S. epidermidis bacterial colonization of biomedical devices. Here, we report on our efforts to (i) isolate AAP, (ii) generate MAbs against AAP, and (iii) determine the efficacy of MAbs to inhibit S. epidermidis biofilm formation. An M7 S. epidermidis mutant, reportedly deficient in AAP expression, was used as a negative control. Postinoculation murine sera, containing polyclonal antibodies against AAP, were able to reduce S. epidermidis biofilm formation by 54%. Select MAbs against AAP were able to reduce S. epidermidis by no more than 66%. Two MAb mixtures, 12C6/12A1 and 3C1/12A1, reduced S. epidermidis accumulation up to 79 and 87%, respectively, significantly more than individual MAbs. Contrary to a previous report, biofilm-deficient S. epidermidis mutant M7 expressed a 200-kDa protein on its cell wall that specifically bound AAP MAbs. Peptide characterization of this M7 protein by microcapillary reversed-phase high-pressure liquid chromatography-nanoelectrospray tandem mass spectrometry resulted in 53% homology with AAP. Ongoing studies will elucidate the dynamic expression of AAP and the M7 200-kDa protein in order to define their roles in biofilm formation.

Binding and orientation of fibronectin on polystyrene surfaces using immobilized bacterial adhesin-related peptides.

Fibronectin (FN) is known to bind to bacteria via high affinity receptors on bacterial surfaces known as adhesins. The binding of bacteria to FN is thought to have a key role in foreign device associated infections. For example, previous studies have indicated that Staphylococcus aureus adhesins bind to the 29 kDa NH(3) terminus end of FN, and thereby promote bacteria adherence to surfaces. Recently, the peptide sequences within the S. aureus adhesin molecule that are responsible for FN binding have been identified. Based on these observations, we hypothesize that functional FN can be bound and specifically oriented on polystyrene surfaces using bacterial adhesin-related (BRP-A) peptide. We further hypothesize that monoclonal antibodies that react with specific epitopes on the FN can be used to quantify both FN binding and orientation on these surfaces. Based on this hypothesis, we initiated a systematic investigation of the binding and orientation of FN on polystyrene surfaces using BRP-A peptide. To test this hypothesis, the binding and orientation of the FN to immobilized BRP-A was quantified using (125)I-FN, and monoclonal antibodies. (125)I-FN was used to quantitate FN binding to peptide-coated polystyrene surfaces. The orientation of bound FN was demonstrated by the use of monoclonal antibodies, which are reactive with the amine (N) or carboxyl (C) termini of the FN. The results of our studies demonstrated that when the BRP-A peptide was used to bind FN to surfaces that: 1. functional FN was bound to the peptide; 2. anti-C terminus antibodies bound to the peptide FN; and 3. only limited binding of anti-N terminus antibodies to peptide-bound FN occurred. We believe that the data that indicate an enhanced binding of anti-C antibodies reactive to anti-N antibodies are a result of the FN binding in an oriented manner with the N termini of FN bound tightly to the BRP-A on the polystyrene surface.

Binding and orientation of fibronectin to silanated glass surfaces using immobilized bacterial adhesin-related peptides.

Previously, we have demonstrated the suitability of bacterial adhesin-related peptides, directly immobilized on polystyrene surfaces, to bind and orient fibronectin (FN). For these studies a method to bind the large protein FN in a desired orientation on a solid substratum was developed which utilizes a bacterial adhesin-related peptide (designated BRP-A), which is known to bind specifically to the NH3-terminus end of FN. Glass substrata was first coated with an amine-terminated silane, followed by streptavidin (SA), which was used as an intermediate tether to bind the biotinylated bacterial adhesin-related peptide. The BRP-A peptide, used for these studies was synthesized with a terminal biotin to assure irreversible coupling of the BRP-A to the streptavidin. The biotinylated BRP-A was next immobilized on the SA-silanated glass surfaces. 125I-FN was used to quantify the amount of FN binding to the (BRP-A):SA-silanated glass surface. Monoclonal antibodies, which react with specific epitopes at either the NH3- or -COOH-termini of FN, were used to quantify the binding and orientation of FN. The results of these studies indicated: (1) FN bound to the BRP-A:SA-silanated glass surface; and (2) the bound FN was oriented such that NH2-terminal region of FN was bound towards the glass surface and the COOH-terminus was oriented away from the glass surface. These studies demonstrate that small peptides can be used to specifically bind and orient large proteins such as FN on the surfaces.

Binding and orientation of fibronectin on surfaces with collagen-related peptides. Student Research Award in the Masters Science Degree Candidate category, 27th annual meeting of the Society for Biomaterials, St. Paul, MN, April 24-29, 2001.

Although fibronectin (FN) has been used in a variety of in vitro studies to enhance cell and bacteria adhesion, relatively little is known about the molecular interactions of FN with surfaces, particularly the interactions that can control the binding, conformation, and functionality of FN on these surfaces. Even less is known about approaches needed to control binding, orientation, and functionality of FN bound on surfaces. To begin to fill this gap in our knowledge, we hypothesized that functional FN can be bound and specifically oriented on polystyrene surfaces with FN-specific collagen-related peptides (CRPs). We further hypothesized that monoclonal antibodies that react with specific epitopes on FN can be used to quantify both FN binding and orientation on these surfaces. On the basis of these hypotheses, we initiated a systematic investigation of the binding and orientation of FN on polystyrene surfaces with CRPs. To bind FN to surfaces, we used two different CRPs: CRP-I (TLQPVYEYMVGV) and CRP-II (TGLPVGVGYVVTVLT). The binding and orientation of the FN molecule to these immobilized CRPs was quantified with (125)I-FN and monoclonal antibodies. Monoclonal antibodies used for this study were reactive with specific regions of the FN molecule, that is, the amino (N) terminus (anti-N antibodies) and carboxyl (C) terminus (anti-C antibodies). The results of our studies demonstrated that although CRP-I and CRP-II could be bound directly to polystyrene, these directly immobilized CRPs failed to bind (125)I-FN . Thus, to facilitate FN binding to the CRPs, we used bovine serum albumin (BSA) as a spacer to physically elevate the CRPs away from the polystyrene surface. Thus, CRP-I and CRP-II were covalently linked to BSA via the N and C termini of each CRP (CRP-I-BSA and CRP-II-BSA). (125)I-CRP-BSAs were all found to bind to equivalent levels on polystyrene (1.60-2.60 microg/cm2). When CRP-BSAs were immobilized on polystyrene, they all successfully bound (125)I-FN in a range of 34-72 ng/cm2 (mean). Using monoclonal antibodies to FN to characterize the orientation of FN bound to the various CRP-BSAs, we demonstrated that (1) FN consistently bound to either CRP-I-BSA or CRP-II-BSA; (2) bound FN reacted significantly more with anti-C antibodies than with anti-N antibodies; and (3) the increased reactivity of bound FN to anti-C antibodies was consistent, whether FN was bound by CRP-I or CRP-II or the CRPs were bound to BSA by the C or N termini. These data demonstrated an enhanced binding of anti-C antibodies to immobilized CRP-BSA relative to anti-N antibodies. We interpreted the data to be the result of FN binding in an oriented fashion with N termini of FN bound tightly to the BSA-polystyrene surface. In this position, the C termini of FN are exposed and available for binding by the anti-C antibodies. Alternatively, in this orientation the N termini of the FN would not be available to bind the anti-N antibodies, thereby explaining the decreased reactivity of the CRP immobilized FN to the anti-N antibodies. These studies not only demonstrate the utility of peptides in binding and orienting large molecular weight proteins such as FN on surfaces but underscore the need for well-characterized reagents (e.g., monomeric/functional FN and antibodies) to specifically bind, orient, and characterize large molecular weight proteins immobilized on various surfaces.

Efficacy of silver-coated fabric to prevent bacterial colonization and subsequent device-based biofilm formation.

Efficacy of silver-coated poly(ethylene terephthalate) to prevent bacterial attachment and subsequent infection was quantified in vitro, in both batch- and flowing-fluid experiments. Kinetic analysis of batch suspended cell cultures of Staphylococcus epidermidis (SE), at various growth-limiting nutrient concentrations, in the absence of any fabric, indicated a maximum culture growth rate constant micro(max) = 0.78 +/- 0.02 h(-1). Batch experiments for Control fabric samples indicated that SE cultures exhibited about the same suspended cell growth rate (0.72 +/- 0.02 h(-1)) as observed in batch suspended cultures without fabric. Suspended SE cultures in the presence of silver-coated fabric grew at a considerably lower rate, 0.15 +/- 0.01 h(-1), indicating the inhibitory effect of Ag(+2) ion released from the fabric. Growth rates of suspended SE cultures were 5-6 times higher in the fluid phase in contact with the Control fabric compared to cultures exposed to silver-coated fabric. Maximum suspended cell concentrations attained at time = 24 h were 1-2 orders of magnitude higher for Control fabrics vs. silver-coated fabric. In all batch colonization experiments, both live and dead SE bacterial cells accumulate on the surfaces of both silver-coated and Control fabrics. Adherent viable SE cells accumulated to 1-2 orders of magnitude more ( approximately 5 x 10(+8) cells/cm(2)) on Control fabric than SE cells on the silver-coated fabric ( approximately 1.1 x 10(+6) cells/cm(2)), respectively. Between 70-95% SE cells on the Control fabric were viable, while on the silver-coated fabric samples, at 24 h, viable cells were less than 10% of the adherent community (i.e., greater than 90% nonviable cells). In flow cell colonization experiments, SE cells accumulated on Control fabric to a maximum adherent cell concentration of 6 x 10(+7) - 8 x 10(+7) cells/cm(2) by 24 h with the proportion of viable cells remaining relatively constant at 76% throughout an experiment. Both noninvasive microscopic enumeration and destructive assays gave the same results for adherent cell numbers. Using silver-coated fabric, total cells numbers (live + dead) reached a level of approximately 1.1 x 10(+7) - 3.0 x 10(+7) cells/cm(2) after about 6 h and remained constant. However, while the proportion of viable cells initially on the surface was 63-75%, this fraction dropped continuously during each experiment to less than 6% viable cells at 24 h. Regardless of the criteria, the number of viable or nonviable cells attached to silver-coated fabric were significantly lower than on Control fabric.

Plasma-deposited membranes for controlled release of antibiotic to prevent bacterial adhesion and biofilm formation.

Bacterial infection on implanted medical devices is a significant clinical problem caused by the adhesion of bacteria to the biomaterial surface followed by biofilm formation and recruitment of other cells lines such as blood platelets, leading to potential thrombosis and thromboembolisms. To minimize biofilm formation and potential device-based infections, a polyurethane (Biospan) matrix was developed to release, in a controlled manner, an antibiotic (ciprofloxacin) locally at the implant interface. One material set consisted of the polyetherurethane (PEU) base matrix radiofrequency glow discharge plasma deposited with triethylene glycol dimethyl ether (triglyme); the other set had an additional coating of poly(butyl methyacrylate) (pBMA). Triglyme served as a nonfouling coating, whereas the pBMA served as a controlled porosity release membrane. The pBMA-coated PEU contained and released ciprofloxacin in a controlled manner. The efficacy of the modified PEU polymers against Pseudomonas aeruginosa suspensions was evaluated under flow conditions in a parallel plate flow cell. Bacterial adhesion and colonization, if any, to the test polymers were examined by direct microscopic image analysis and corroborated with destructive sampling, followed by direct cell counting. The rate of initial bacterial cell adhesion to triglyme-coated PEU was 0. 77%, and to the pBMA-coated PEU releasing ciprofloxacin was 6% of the observed adhesion rates for the control PEU. However, the rate of adherent cell accumulation due to cell growth and replication was approximately the same for the triglyme-coated PEU and the PEU controls, but was zero for the pBMA-coated PEU releasing ciprofloxacin.

Design of infection-resistant antibiotic-releasing polymers: I. Fabrication and formulation.

Biomaterials-related infections are often observed with prosthetic implants and in many cases result in the failure of the devices. To design a biomedically useful polymer that is intrinsically infection-resistant, we have developed a ciprofloxacin-loaded polyurethane (PU) matrix that releases antibiotic locally at the implant surface, thereby minimizing bacterial accumulation. We report here the methods of fabrication and formulation for making such antibiotic-loaded devices, as well as evidence of their bactericidal properties. Specifically, various pore-forming agents and drug loadings were examined. An optimum formulation consisting of BIOSPAN PU, poly(ethylene glycol) and ciprofloxacin offered the longest effective period of sustained release (5 days). The bactericidal efficacy of the released ciprofloxacin against Pseudomonas aeruginosa (PA) was four times that of the control PU without antibiotics. This bactericidal efficiency was due to an increase in the PA detachment from the surface. These observations suggested that the released ciprofloxacin was biologically active in preventing the bacteria from permanently adhering to the substratum, and thus decreasing the possibility of biofilm-related infection.

Mobilization of broad host range plasmid from Pseudomonas putida to established biofilm of Bacillus azotoformans. I. Experiments.

A strain of Pseudomonas putida harboring plasmids RK2 and pDLB101 was exposed to a pure culture biofilm of Bacillus azotoformans grown in a rotating annular reactor under three different concentrations of the limiting nutrient, succinate. Experimental results demonstrated that the broad host range RSF1010 derivative pDLB101 was transferred to and expressed by B. azotoformans. At the lower concentrations, donor mediated plasmid transfer increased with increasing nutrient levels, but the highest nutrient concentration yielded the lowest rate of donor to recipient plasmid transfer. For transconjugant initiated transfer, the rate of transfer increased with increasing nutrient concentrations for all cases. At the lower nutrient concentrations, the frequency of plasmid transfer was higher between donors and recipients than between transconjugants and recipients. The reverse was true at the highest succinate concentration. The rates and frequencies of plasmid transfer by mobilization were compared to gene exchange by retrotransfer. The initial rate of retrotransfer was slower than mobilization, but then increased dramatically. Retrotransfer produced a plasmid transfer frequency more than an order of magnitude higher than simple mobilization.

Activity and stability of a recombinant plasmid-borne TCE degradative pathway in suspended cultures.

The retention and expression of the plasmid-borne, TCE degradative toluene-ortho-monooxygenase (TOM) pathway in suspended continuous cultures of transconjugant Burkholderia cepacia 17616 (TOM31c) were studied. Acetate growth and TCE degradation kinetics for the transconjugant host are described and utilized in a plasmid loss model. Plasmid maintenance did not have a significant effect on the growth rate of the transconjugant. Both plasmid-bearing and plasmid-free strains followed Andrews inhibition growth kinetics when grown on acetate and had maximum growth rates of 0.22 h-1. The transconjugant was capable of degrading TCE at a maximum rate of 9.7 nmol TCE/min. mg protein, which is comparable to the rates found for the original plasmid host, Burkholderia cepacia PR131 (TOM31c). The specific activity of the TOM pathway was found to be a linear function of growth rate. Plasmid maintenance was studied at three different growth rates: 0.17/h, 0.1/h, and 0.065/h. Plasmid maintenance was found to be a function of growth rate, with the probability of loss ranging from 0.027 at a growth rate of 0.065/h to 0.034 at a growth rate 0.17/h.

Mobilization of broad host range plasmid from Pseudomonas putida to established biofilm of Bacillus azotoformans. II. Modeling.

A strain of Pseudomonas putida that harbors plasmids RK2 and pDLB101 was exposed to a pure culture biofilm of Bacillus azotoformans grown in a rotating annular reactor. Transfer of the RK2 mobilizable pDLB101 plasmid to B. azotoformans was monitored over a 4-day period. Experimental results demonstrated that the broad host range, RSF1010 derivative pDLB101 was transferred to and expressed by B. azotoformans. In the companion article to this work, the rate of plasmid transfer was quantified as a function of the limiting nutrient, succinate, and as a function of the mechanism of transfer. A biofilm process simulation program (AQUASIM) was modified to analyze resultant experimental data. Although the AQUASIM package was not designed to simulate or predict genetic events in biofilms, modification of the rate process dynamics allowed successful modeling of plasmid transfer. For the narrow range of substrate concentrations used in these experiments, nutrient level had only a slight effect on the rate and extent of plasmid transfer in biofilms. However, further simulations using AQUASIM revealed that under nutrient poor conditions, the number of transconjugants appearing in the biofilm was limited.

Local macromolecule diffusion coefficients in structurally non-uniform bacterial biofilms using fluorescence recovery after photobleaching (FRAP).

Pure culture Pseudomonas putida biofilms were cultivated under controlled conditions to a desired overall biofilm thickness, then employed within classical half-cell diffusion chambers to estimate, from transient solute concentrations, the effective diffusion coefficient for several macromolecules of increasing molecular weight and molecular complexity. Results of traditional half-cell studies were found to be erroneous due to the existence of microscopic water channels or crevasses that perforate the polysaccharidic gel matrix of the biofilm, sometimes completely to the supporting substratum. Thus, half-cell devices measure a composite transfer coefficient that may overestimate the true, local flux of solutes in the biofilm polysaccharide gel matrix. An alternative analytical technique was refined to determine the local diffusion coefficients on a micro-scale to avoid the errors created by the biofilm architectural irregularities. This technique is based upon the Fluorescence Return After Photobleaching (FRAP), which allows image analysis observation of the transport of fluorescently labeled macromolecules as they migrate into a micro-scale photobleached zone. The technique can be computerized and allows one to map the local diffusion coefficients of various solute molecules at different horizontal planes and depths in a biofilm. These mappings also indirectly indicate the distribution of water channels in the biofilm, which was corroborated independently by direct microscopic observation of the settling of fluorescently-labeled latex spheres within the biofilm. Fluorescence return after photobleaching results indicate a significant reduction in the solute transport coefficients in biofilm polymer gel vs. the same value in water, with the reduction being dependent on solute molecule size and shape.

Toluene degradation kinetics for planktonic and biofilm-grown cells of Pseudomonas putida 54G.

Toluene degradation kinetics by biofilm and planktonic cells of Pseudomonas putida 54G were compared in this study. Batch degradation of (14)C toluene was used to evaluate kinetic parameters for planktonic cells. The kinetic parameters determined for toluene degradation were: specific growth rate, micro(max) = 10.08 +/- 1.2/day; half-saturation constant, K(S) = 3.98 +/- 1.28 mg/L; substrate inhibition constant, K(I) = 42.78 +/- 3.87 mg/L. Biofilm cells, grown on ceramic rings in vapor phase bioreactors, were removed and suspended in batch cultures to calculate (14)C toluene degradation rates. Specific activities measured for planktonic and biofilm cells were similar based on toluene degrading cells and total biomass. Long-term toluene exposure reduced specific activities that were based on total biomass for both biofilm and planktonic cells. These results suggest that long-term toluene exposure caused a large portion of the biomass to become inactive, even though the biofilm was not substrate limited. Conversely, specific activities based on numbers of toluene-culturable cells were comparable for both biofilm and planktonically grown cultures. Planktonic cell kinetics are often used in bioreactor models to model substrate degradation and growth of bacteria in biofilms, a procedure we found to be appropriate for this organism. For superior bioreactor design, however, changes in cellular activity that occur during biofilm development should be investigated under conditions relevant to reactor operation before predictive models for bioreactor systems are developed.

Bacterial infection of biomaterials. Experimental protocol for in vitro adhesion studies.

Some of the more common reactor systems and novel diagnostic tools employed in the study of bacterial cell adhesion and biofilm formation have been described. Sampling and experimental requirements are shown to greatly influence the design and construction of a biofilm reactor. As analytical techniques evolve, the capability to non-invasively follow the development of biofilms and to assess the attached cell reactivity has increased. Both non-invasive and invasive diagnostic methods affect the type and design of biofilm flow reactor with both types of analyses providing complementary information on biofilm processes. To correctly interpret the contribution of a specific rate process to the net accumulation of cells at a substratum, one requires a reactor system devoid of any mass transfer limitations and a process analysis approach to allow for the correct collection and analysis of data.

Xenobiotic biodegradation test using attached bacteria in synthetic seawater.

The aerobic biodegradability of aniline, used as reference chemical, has been performed in synthetic seawater with attached biomass in a continuously fed reactor (biofilm chemostat reactor, BCR). Marine bacteria inocula came from local marine fish aquarium filters to limit the geographic and seasonal variations in quality. A pretreatment of these inocula combining 5-microns filtration and centrifugation was used to concentrate bacteria and remove organic carbon contamination of the test. The performances of the BCR were tested in comparison with simple shake flask tests. Among the different variables tested, the ratio S0/X0 (initial concentration of xenobiotic to initial density of the inoculum), the presence of dissolved oxygen, and the hydraulic residence time appear to be the key parameters controlling the length of the biodegradation process. On the other hand, the addition of a cosubstrate (easily biodegradable compound) does not provide advantages. Thus, marine biofilm chemostat reactors with a high density of attached bacteria (around 10(7) cells cm-2) and fed with synthetic seawater plus nitrogen provide good tools for screening biodegradability of chemicals in the marine environment.