The influence of calcium on biofilm processes by Mukesh Harilal Turakhia A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Chemical Engineering Montana State University © Copyright by Mukesh Harilal Turakhia (1986) Abstract: Bacteria exhibit a tendency for adsorbing to and colonizing surfaces which are submerged in aquatic environments. Adsorption is mediated by extracellular polymeric material which is formed by the bacteria and extends from the cell to the attachment surface. The attached cells reproduce and form additional extracellular polymer increasing the mass of the deposit. The cellular-extracellular matrix is termed a bipfilm. The purpose of this study was to investigate the effect of calcium on cellular reproduction and extracellular polymer formation by Pseudomonas aeruginosa in a biofilm. Experiments were conducted with a pure culture of Ps. aeruginosa using fixed film bioreactors with glucose serving as the limiting nutrient. Results indicate calcium increases the rate and extent of cellular carbon accumulation at the surface. However, there was no effect of calcium on the amount of polymer carbon accumulated on the surface. Results also suggest that free calcium (or calcium-assisted ligands) is essential to the structural integrity of the biofilm. The energy required for biochemical conversion of glucose into biomass by suspended or immobilized culture of Ps. aeruginosa was constant and was independent of time, biomass concentration, specific cellular growth rate, and calcium concentration in the medium.  THE INFLUENCE OF CALCIUM ON BIOFILM PROCESSES by Mukesh Harilal Turakhia A thesis submitted in partial fulfillment, of the requirements for the degree of Doctor of Philosophy in ■ ' Chemical Engineering MONTANA STATE UNIVERSITY Bozeman, Montana March 1986. ii j>37% T%4 / o L v APPROVAL of a thesis submitted by Mukesh Harilal Turakhia This thesis has been read by each member of the thesis committee and been found to be satisfactory regarding content, English usage, format, citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies. Date Chairperson, Graduate Committee Approved for Major Department Date Approved for College of Graduate Studies /9_____L'Tuyc&k i Date iv ACKNOWLEDGMENTS I wish to express my appreciation to the following: Bill Characklis for his support, guidance, and encouragement which made my graduate educational experience productive and enjoyable. Keith Cooksey, Gordon McFeters, and Dan Goodman for their advice and critical comments relating to my thesis. The "slime gang": Rune, Andy, Maarten, Bjorn, Pam, Nick, Frank, Joe, Chris, and Rich for being an interesting group with whom to work and interact. Lyman Fellows, Gordon Williamson, and Stuart Aasgaard for their help in maintenance of various equipment. The staff and students of the Chemical Engineering, Civil Engineering, and Microbiology Departments whom I have not mentioned but have directly or indirectly contributed to this project. I-Hsing Tsao for biofilm elemental analysis, Robert Kultgen for drafting some of the figures, and Wanda Myers for typing this thesis. Naval Post Graduate Foundation for providing partial financial support through The Carl E. Menneken Fellowship. Montana State University Engineering Experiment Station, National Science Foundation, and Office of Naval Research for partial financial support. V TABLE OF CONTENTS LIST OF TABLES ............................ LIST OF FIGURES ............................ ABSTRACT . . . . .......................... INTRODUCTION .............................. Research Goal and Objectives . ... . . . LITERATURE REVIEW................... . . . Biofilm Formation: A Process Analysis . . Organic Adsorption .................... Transport of Microbial Cells .......... Microbial Adsorption .................. Microbial Transformation .............. Biofilm Detachment .................... Organism.......... .................... MATHEMATICAL DESCRIPTION OF THE SYSTEM . . . Chemostat Equation .................... Annular Reactor Equations .............. EXPERIMENTAL APPARATUS AND METHODS . . . . . Experimental Apparatus ................ Experimental Procedure .......... .. Experimental Start-Up .............. • • Sampling .............................. Analytical Methods ........ . . . . . . RESULTS .................................... Influence of Calcium on Biofilm Formation Effect of Calcium on Accumulation Rates . Effect of Calcium on Process Rates . . . Batch Growth Experiments .............. Page vii I x xiii 1 2 4 14 16 18 21 22 29 29 40 44 46 47 53 53 55 64 91 m m vo CO vi TABLE OF CONTENTS (Continued) Page DISCUSSION...................................................... 99 Elementary Composition of Microorganisms .................... 99 Biofilm Calcium.............................................. 103 Specific Substrate Removal Rate ........................ 105 Specific Oxygen Removal R a t e ........ ..................... 107 Specific Cellular Growth R a t e .......................... 109 Cellular Detachment R a t e ................................. 113 Specific Polymer Production Rate ............................ 115 Stoichiometric Coefficient.................. '............. 117 Substrate Diffusion........................... 119 Calcium and Biofilm Structure....... ......................120 CONCLUSIONS . . . ................................................ 123 LITERATURE C I T E D ................................................ I25 NOMENCLATURE...................................... 134 APPENDICES...................................................... 137 Appendix A - Raw Data: Annular Reactor Experiments .......... 138 Appendix B - Raw Data: Tubular Reactor Detachment Experiments.............................................. 149 Appendix C - Raw Data: Annular Reactor Detachment • Experiments.............................................. 133 Appendix D - Effluent Glucose and Oxygen Data (AR Experiments)........................................ 104 Appendix E - Raw Data: Batch Growth Experiments ............... 173 Appendix F - Oxygen Diffusion Studies . ...................... 179 Appendix G - Batch Growth Experiments: Effect of EGTA . . . . 183 Appendix H - The Influence of Calcium on Microbial Adsorption ................................. 187 vii LIST OF TABLES Table Page 1 Relevant Characteristics and Dimensions of Chemostat. 31 2 Composition of Growth Medium for Chemostat and Annular Reactor Experiments. 32 3 Composition of Micronutrients Calculated for a Glucose Concentration of 1000 g m~3. 33 4 Relevant Characteristics and Dimensions of the Annular Reactor. 35 5 Growth Medium Composition for Tubular Reactor Experiments. 39 6 Composition of Batch Growth Medium for BOD Respirometer Experiments. 41 7 Summary of Experimental Conditions of Annular Reactor Experiments, 54 8 Composition of Biofilm (Dry Weight Basis). 62 9 Summary of Annular Reactor Detachment Experiments. 75 10 Experimental Conditions and Summary of Tubular Reactor Results. 76 11 Change in Effluent Glucose Carbon Concentration as a Result of Chelant Addition (AR Experiment 4). 87 12 Biofilm Calcium Before and After Chelant (EGTA) Addition. 90 13 Experimental Conditions and Summary of Batch Growth Experimental Results. 91 14 Chemical Composition of Various Biofilms and Suspended Cultures. 100 15 Elemental Composition of Biofilm (Ash Free Basis). 101 16 Elemental Composition for Biomass of Various Sources. 102 viii LIST OF. TABLES (Continued) Table Page 17 Effect of EGTA (of Free Calcium) on Maximum Specific Growth Rate of Pseudomonas aeruginosa. 112 18 Experimental Results from Adsorption Studies Using Pseudomonas 224S. 122 19-27 Raw Data Annular Reactor Experiments. 140 28-29 Raw Data Tubular Reactor Detachment Experiments. 151 30-38 Raw Data Annular Reactor Detachment Experiments. 155 39-46 Effluent Glucose and Oxygen Data from AR Experiments. 165 47-49 Oxygen Uptake Data Batch Experiments. 174 50 Experimental Data from Oxygen Transport Studies. 182 51 Composition of Batch Growth Medium. 186 ix LIST OF FIGURES Figure ^aBe 1 A simplified diagram of the annular reactor. 19 2 A simplified diagram of the chemostat. 20 3 The annular reactor (AR) was operated as a continuous flow stirred tank reactor (CFSTR). The insert describes the local environment in the reactor. The mathematical model was based on this conceptual model. 23 4 A schematic diagram of BOD respirometer. 30 5 Flow diagram of the annular reactor system. 37. 6 Simplified flow diagram of the tubular reactor. 38 7 Hierarchy of analytical procedures. 48 8 Progression of biofilm mass as a function of time for three different calcium concentrations. Lines drawn by observation. 56 9 Change in biofilm mass (140 h) as a function of calcium in the dilution water. Error bars represent standard deviation of measurements of all experiments with same calcium concentration in the dilution water. 57 10 Progression of biofilm cellular carbon as a function of time for three different calcium concentrations. Curves represent time smoothed data. Parameters for time smoothed data were obtained using BMDP3R nonlinear regression. 58 11 Change in biofilm cellular and polymer carbon (140 h) as a function of calcium in the dilution water. Error bars represent standard deviation of measurements of all experiments with same calcium concentration in the dilution water. 60 12 Change in biofilm calcium as a function of calcium in the dilution water. 61 X LIST OF FIGURES (Continued) Figure ^aSe 13 Change in effluent glucose (carbon equivalents) as a function of time. Curve represents time smoothed data (Experiment 6, AR #8). 63 14 Change in effluent oxygen concentration as a function of time. Line drawn by observation (Experiment 6; AR#8). 65 15 Change in specific substrate removal rate as a function of time with calcium in the dilution water as a parameter. Curves represent time smoothed data. 67 16 Change in specific oxygen removal rate as a function of time with calcium in the dilution water as a parameter. Curves represent time smoothed data. 69 17 Change in biofilm specific cellular growth rate as a function of time. Line drawn by observation. 70 18 Change in biofilm specific cellular detachment rate as a function of time. Line drawn by observation. 71 19 Change in biofilm specific polymer production rate as a function of time. Line drawn by observation. 73 20 Change in biofilm specific polymer detachment rate as a function of time. Line drawn by observation. 74 21 Response of a biofouled annular reactor (Experiment 6) to the addition of I mM EGTA. Lines drawn by observation. 78 22 Response of a biofouled annular reactor (Experiment 2; AR #8) to the addition of 0.5 mM EGTA. Lines drawn by observation. 79 23 Response of a biofouled tubular reactor (Experiment #1) to the addition of I mM EGTA. Lines drawn by observation. 80 24 Response of a biofouled tubular reactor (Experiment #2) to the addition of I mM EGTA. Lines drawn by observation. 81 xi Figure 25 26 27 28 29 30 31 32 33 34 LIST OF FIGURES (Continued) Change in suspended cell concentration in the effluent of the annular reactor (Experiment 2; AR #8) after EGTA (I mM) addition. Line drawn by observation. Error bars represent standard deviation of measure^ ments of the same sample. Change in total carbohydrate (as glucose equivalents) in the effluent of the annular reactor (Experiment 2; AR #8) as a result of EGTA (I mM) addition. Line drawn by observation. • / Change in suspended solids and suspended cell mass in the effluent due to the addition of I mM EGTA. Lines drawn by observation. Error bars based on standard deviation from epifluorescent cell counts. Change in total carbohydrate (as glucose equivalents) in the effluent of tubular reactor (Experiment #1) after the addition of I mM EGTA. -Line drawn by observation. i Change in oxygen uptake as a function of time for two different calcium concentrations (Batch Experiment //I). Change in oxygen uptake as a function of time for two different calcium concentrations (Batch Experiment #2). Change in oxygen uptake as a function of time for two different calcium concentrations (Batch Experiment #3). Change in oxygen uptake as a function of time on a semi-log coordinate for different calcium concen­ trations (Batch Experiment #1). Change in specific substrate removal rate as a function of biofilm specific cellular growth rate. The straight line represents the best-fit described by.Equation 19. Change in specific oxygen removal rate as a function of biofilm specific cellular growth rate. The straight line represents the best-fit described by Equation 21. Page 84 85 88 89 93 94 95 96 106 108 xii Figure 35 36 37 38 LIST OF FIGURES (Continued) Plot of biofilm specific cellular growth rate as a function of glucose carbon concentration. The shaded portion represents the 95 percent confidence interval of the Monad equation based on parameters (y and K^) estimated by Trulear (1983) for Psv aeruginosa in suspended culture. Ratio of specific cellular detachment rate to specific cellular production rate as a function of time. Lines drawn by observation. Plot of specific polymer production rate as a function of biofilm specific cellular growth rate. The straight line represents the best-fit described by Equation 9. Change in specific substrate removal rate as a function of specific oxygen removal rate. The straight line represents the best linear fit. Page H O 114 116 118 xiii ABSTRACT Bacteria exhibit a tendency for adsorbing to and colonizing surfaces which are submerged in aquatic environments. Adsorption is mediated by extracellular polymeric material which is formed by the bacteria and extends from the cell to the attachment surface. The attached cells reproduce and form additional extracellular polymer increasing the mass of the deposit. The cellular-extracellular matrix is termed a bipfilm. , The purpose of this study was to investigate the effect of calcium on cellular reproduction and extracellular polymer formation by Pseudomonas aeruginosa in a biofilm. Experiments were conducted with a pure culture of Ps_. aeruginosa using fixed film bioreactors with glucose serving as the limiting nutrient. Results indicate calcium increases the rate and extent of cellular carbon accumulation at the surface. However, there was no effect of calcium on the amount of polymer carbon accumulated on the surface. Results also suggest that free calcium (or calcium-assisted ligands) is essential to the structural integrity of the biofilm. The energy required for biochemical conversion of glucose into biomass by suspended or immobilized culture of Ps_. aeruginosa was constant and was independent of time, biomass concentration, specific cellular growth rate, and calcium concentration in the medium. I INTRODUCTION Microorganisms, primarily bacteria, exhibit a tendency for adsorbing to and colonizing surfaces which are submerged in aquatic environments. The immobilized cells grow, reproduce, and produce extracellular polymeric substances (EPS) which frequently extend from the cell, forming a tangled mass of fibers lending structure to the entire assemblage which shall be termed a biofilm. Biofilms serve beneficial purposes in natural environments and in some modulated systems. For example, biofilms are responsible for removing organic or inorganic "contaminants" from natural streams and in wastewater treatment processes (e.g., trickling filters and rotating biological contactors). Biofilms can, however, impair the performance of process equipment. They can impede the flow of heat across the surface, increase fluid frictional resistance at the surface, and increase the corrosion rate at the surface. Fouling of heat exchange equipment was estimated to cost the United States billions of dollars annually (Lund and Sandu, 1981). Most studies on the effect of dissolved constituents on biofilm formation have been limited to the effect of organic constituents. There is little or no information on the effect of dissolved inorganic constituents. There are a number of inorganic components which can have an effect. However, the scope of this study was limited to the effect of calcium on biofilm formation. 2 The presence of calcium in the microbial growth medium has been shown to (I) influence microbial adsorption to the solid substratum, (2) be required for cellular growth and reproduction, (3) influence the composition of EPS. Published reports on the influence of calcium on the above processes is contradictory and the majority of the studies were conducted in quiescent conditions where the cells are not subject to shear stress. Understanding the role of calcium in biofilm proces­ ses may be useful in ecosystem analysis, control of biofouling in heat exchangers and/or pipelines, operation of fixed film biological waste- water treatment processes and increasing the rate, or strength of cell immobilization in a biofilm reactor for biotechnology applications. The major emphasis of this study was to determine the influence of calcium on (I) cellular reproduction and polymer formation in a bio­ film, and (2) the cohesive strength of biofilm. Pseudomonas aeruginosa was used as a test organism because this organism has been extensively studied both in continuous flow stirred tank reactors, i.e., chemo- stats, and in biofilm reactors. Specific cellular growth rates and EPS formation rates are known under defined experimental conditions in chemostat and biofilm environments (Robinson et al., 1984; Bakke et al., 1984). Research Goal and Objectives The goal of this research was to determine.the role of calcium in the formation and maintenance of a biofilm. To accomplish this goal, the following objectives were established: 3 1. Determine the influence of calcium on cellular reproduction and extracellular polymer formation by Pseudomonas aeruginosa in a biofilm. 2. Determine the influence of calcium on biofilm cohesiveness. 3. Determine the influence of calcium on the net accumulation rate of biofilm resulting from processes and characteristics in Objectives (I) and (2). 4 LITERATURE REVIEW Biofilm Formation: A Process Analysis The adsorption of bacteria is a.general phenomenon encountered in natural environments with important ecological implications. Bacterial adsorption to surface offers advantages . in terms of nutrient avail­ ability, particularly in fast flowing and nutrient deficient habitats. The adsorbed cells reproduce and form extracellular polymers leading to the formation of a biofilm. Accumulation of biofilm at the surface is the net result of the following fundamental processes (Characklis, 1981): 1. Adsorption of organic molecules to the surface forming a conditioned surface. 2. Transport of microbial cells to the conditioned surface. 3. Microbial adsorption to the conditioned surface. 4. Microbial transformation (growth, reproduction, etc.) at the surface resulting in the formation of biofilm. 5. Partial detachment of biofilm due to fluid shear stress. Biofilm formation is not a sequence of the above rate processes occurring individually but rather the net result of these processes occurring simultaneously. At specific times in the overall develop­ ment, certain rate processes contribute more than the others. This literature review will focus on the influence of calcium on the processes involved in the formation of biofilm. 5 . Organic Adsorption Adsorption of an organic monolayer occurs within minutes of exposure of an initially "clean" surface to an aqueous environment containing dissolved organics, microorganisms, and nutrients. This adsorption changes the properties of the wetted surface and actually conditions the surface for subsequent attachment and colonization (Loeb and Neihof, 1975; Baler and Depalma, 1977). These conditioning films have been investigated by various means. Baier and various co-workers have characterized these acquired films as negatively charged (poly­ anionic) polysaccharides or glycoproteins (Baier, 1980, Barer and Weiss, 1975; Marshall, 1979). There appears to be no evidence, however, that microorganisms can only attach to conditioned surfaces. Also, little is known regarding the influence of calcium on organic adsorption. Transport of Microbial Cells Microbial cells (0.5 - 10.0 pm) can be transported from the bulk fluid to the wetted surface by several mechanisms, including the following: diffusion (Brownian), gravity, thermophoresis, taxis,, and fluid dynamic forces (inertia, drag, drainage, and downsweeps). In general, the transport of microbial cells from the bulk fluid to the wetted surface depends on fluid flow conditions and is not known to be influenced by the presence of calcium in the bulk fluid. 6 Microbial Adsorption Once bacterial cells have been transported to the wetted surface, two types of adsorption are possible; reversible and irreversible (Marshall et al., 1971; Zobell, 1943). Reversible adsorption is characterized by an initially weak adsorption of a cell which still exhibits Brownian motion and is readily removed by mild rinsing. Conversely, 'irreversible adsorption is a permanent bonding to the surface, usually aided by the production of EPS (Fletcher, 1980). Cells attached in this way can only be removed by rather severe mechanical or chemical treatment. Calcium and Microbial Adsorption The role of cations in the adsorption of a cell to the substratum is presently unknown. Roux (1894) reported the necessity for divalent cations, notably Ca2+, in cellular adsorption. Calcium has been shown to be necessary for adsorption of. aquatic bacteria (Marshall et al., 1971; Fletcher and Floodgate, 1973; Stanley, 1983) and marine diatoms (Cooksey, 1981). For example, a marine pseudomonad would not irrevers­ ibly adsorb in the absence of Ca and Mg , but would adsorb when either of the cations were present (Marshall et al., 1971). Stanley (1983) observed that Ps. aeruginosa adsorbed poorly in distilled water with adsorption increasing as calcium chloride concentration was increased to 10 mM. Fletcher (1980) observed the influence of cations on the adsorption process by "chemically treating" free-living cells and 7 observing any influence on their subsequent adsorption to surfaces. The adsorption of a marine pseudomonad (Fletcher, 1980) was inhibited by the presence of EDTA (ethylenediaminetetra-acetic acid) suggesting that the chelant removed surface-bound divalent cations conceivably involved in intercellular-ionic bridging. Fletcher (1980) noted that lanthanum decreased bacterial adsorption and postulated that lanthanum prevented adsorption through interaction with and subsequent denatura- tion of EPS. Lanthanum is known to inhibit calcium transport into cells, and to displace calcium from cellular membranes (Weiss, 1974), so that the effect observed by Fletcher (1980) may have been related to the diminution of the flux of calcium to the intracellular space. Divalent cations, calcium in particular, have been shown to influence microbial adsorption. However, the role of cations in the adsorption process is much disputed. It has been suggested that cations influence adsorption (I) by directly influencing cell physiology or membrane permeability (Drapeau and MacLeod, 1965), and (2) directly by their accumulation at the cell surface where they mediate the formation of the electric double layer (Shaw, 1970). Moscona (1968) suggested that the removal and/or absence of divalent cations inhibits adsorption via calcium sensitive ligands. It has also been suggested that divalent cations, especially calcium, can form bridges between negatively charged substrata and microorganisms, can stabilize the structure of EPS (Fletcher and Floodgate, 1973), or cause precipitation of EPS in the space between a cell and a substratum (Rutter, 1980). Further evidence for the involvement of calcium in the adsorption process comes from the use of complexing. agents EDTA 8 (Fletcher, 1980) and EGTA (Appendix H) „ The chelant did not remove cells irreversibly adsorbed (ethylene glycol-bis(8-aminoethyl ether)-N, N-tetraacetic acid) to the surface (Fletcher, 1980; Appendix H) suggesting that calcium was not involved in the adsorption of cells to the substratum. Microbial Transformation The attached microorganisms assimilate nutrients, reproduce, and form extracellular polymers. The combined result of these processes is the formation of biofilm. The characteristic of the biofilm accumu­ lated will depend on the microbial species, the polymers produced, and the environmental conditions. Biofilm studies thus far (Kornegay and Andrews, 1967; Lamotta, 1967; Zelver, 1979; Trulear and Characklis, 1982) relied on a relative­ ly unstructured approach to the analysis of biomass components. The biotic component was generally characterized only in terms of cell number and cell mass with little attention to the physiological state of the organisms, although there have been some limited attempts at distinguishing between reproduction and polymer formation (Trulear, 1983; Bakke et al., 1984). Trulear (1983) and Bakke et al. (1984) have used process analysis techniques in experimental biofilm reactors to quantify the fundamental rate processes within a biofilm at steady state. Their results suggest the following: I. Psv aeruginosa does not behave differently in biofilms than in suspension at steady state. The biofilm activity measurements in 9 this study were made iji situ, and without significant diffusional resistance. 2. Intrinsic rate and stoichiometric coefficients derived in the ,chemostat can, therefore, be used to describe steady state biofilm processes. Process analysis techniques can be useful in determining the effect of calcium on cellular reproduction and polymer formation. The effect of calcium on microbial transformation within a biofilm can also be inferred from more easily observed rate processes such as substrate consumption and oxygen consumption. However, the observed rate proces­ ses are not a sufficient criterion for comparing microbial activity under different experimental conditions because they are the net result of several fundamental processes. Calcium and Microbial Growth Calcium plays a vital function inside the cell. The concept of calcium as intracellular messenger and/or regulator was proposed 30-40 years ago (Campbell, 1983). For example, Hojeberg and Rydstrom (1977) suggested that calcium is a potent positive effector of nicotinamide nucleotide transhydrogenase in Ps. aeruginosa. Several extracellular degradative enzymes in both eucaryotes and procaryotes require calcium for stability and/or maximal activity (Campbell, 1983). Calcium is required for growth and function of many bacterial species (Campbell, 1983, Weinberg, 1977). Marshall et al. (1971) reported that omission of calcium and magnesium from artificial sea water prevented growth and polymer production of Pseudomonas R3. 10 Shooter and Wyatt (1955) investigated the mineral requirements of Staphylococcus pyogenes and found that calcium and magnesium were needed for growth. Kenward et al. (1979) reported that inclusion of calcium and/or magnesium in the media had no effect on the exponential growth rate (0.66 h"1) of Ps. aeruginosa. Calcium was shown to be necessary for the growth of marine Bdellovibrio sp. (Huang and Staff, 1973; Bell and Lantham, 1975). Calcium appears to be exclusively extracellular and is not accumu­ lated by a normal growing cell (Silver, 1977; Belliveau and Lanyi, 1978; Wacker and William, 1968) . A special situation of calcium accumulation occurs during unusual conditions such as bacterial sporu- lation. Similarly, a number of different major ions (e.g., magnesium, iron, sodium, and potassium) were shown to be required for growth (Shankar and Bard, 1952; Weinberg, 1977; Shooter and Wyatt, 1955). Very little is known regarding the effect of calcium on specific cellular growth rate and/or polymer production. A clear-cut require­ ment of calcium for growth in microorganisms has rarely been demon­ strated (Wyatt, 1961, 1964; Wyatt et al., 1962; Hunter, 1972). This is due to the fact that it is very difficult to reduce the concentration of free calcium below I pM. The concentration of the free calcium can be lowered by the addition of calcium-specific chelant (EGTA). Turakhia (1984) was able to grow Ps. aeruginosa in the presence of 0.006 M EGTA (free calcium in the media was approximately 10 M) . His results (Appendix G) showed that either EGTA or free calcium affected the maximum specific growth rate of Ps^ aeruginosa. 11 Bacterial EPS The formation of extracellular polymer has long been recognized as an important process in the metabolism of many dispersed and immobilized bacteria. Traditionally, two types of extracellular polymer have been distinguished depending on the spatial association of the polymer with the cell (Brock, 1979). Extracellular polymer which remains in a rather compact layer attached to the cell is referred to as a capsule. Conversely, extracellular polymer which does not exhibit a close association with the cell and can exist as a rather dispersed accumulation is referred to as a slime layer. The capsule-slime component of biofilms is termed extracellular polymeric substances (EPS) because little is known about its composition. EPS Characterization. Numerous microorganisms produce exopoly­ saccharides, i.e. polysaccharide found outside the cell wall, either attached to the cell in the form of capsules or secreted into the extracellular environment in the form of slime. Such polymers vary considerably in their chemical structures. There are many qualitative analyses of bacterial EPS, usually considered to be carbohydrate with acidic groups (Corpe et al., 1976; Fletcher and Floodgate, 1973), amino groups (Baier, 1975), and sometimes associated with proteins (Corpe et hi., 1976). A variety of chemical structures is represented in the polysaccharides synthesized by bacteria (Sutherland, 1982). Some components such as D-glucose,■D-mannose, D-galactose, and D-glucuronic acid occur very frequently; others such as L-rhamnose, L-fucose, D-mannuronic acid, and D-guluronic acid are slightly less common. 12 Ps. nerufttnosa EPS. Pseudomonas aerugjnosa was the test organism for this experimental program. The literature on the composition of slime produced by dispersed culture of Ps_. aeruginosa is contradictory. Eagon (1956) reported that Psv aeruginosa produces slime, which consists largely of mannans, but no uronic acid or amino sugars were detected. Later, Eagon (1962) showed that, in addition to mannose, which accounted for 50% of the material, the slime contained appreci­ able amounts of nucleic acid (mostly DNA), and small amounts of proteins. Linker and Jones (1964) showed the production, by a pathor- genic Pseudomonas organism, of a polysaccharide very similar to alginic acid, a polyuronide usually obtained from sea weed. Both mannuronic and small amounts of guluronic acid appear to be present. Carlson and Mathews (1966) have reported that Ps_. aeruginosa slime is a polymer composed of uronic acids. Brown et al. (1969) reported the slime produced by eight strains of Psv aeruginosa (stationary phase) to be qualitatively the same. The slime was shown to be predominantly polysaccharide (mainly glucose with smaller amount of mannose) with some nucleic acids material and a small amount of protein. The extracellular polymers have been shown to be involved in the selective accumulation of ions in many gram negative bacteria (Galanos et al., 1977; Leive, 1974). Buckmire (1983) observed preferential adsorption of Ca, K, P, and S (2-4 times greater than in growth medium) on individual cell and extracellular components, indicating the role of EPS (and possibly associated macromolecules) in the adsorption of ions. The lipopolysaccharide of gram-negative bacteria contains a number of potential cation-binding sites (Schindler and Osborn, 1979; Galanos et 13 al., 1977) having a high affinity for calcium and magnesium. Outside the cell, calcium can bind to carboxylate and sulfate groups of many polysaccharides, many of which are also linked to proteins (Levine and William, 1984). Calcium and EPS Formation Wilkinson and Stark (1956) observed that calcium, magnesium, and potassium stimulated polysaccharide production by Enterobacter aerogenes. Linker and Evans (1976) observed that the composition of Pseudomonas aeruginosa alginate was not influenced by different calcium levels in the medium. However, the composition of Az_. vineland ii (Larsen and Haug, 1971) was affected by the calcium concentration in the medium and postulated that mannuroni.c acid residues were epimerised to guluronic residue by an extracellular enzyme dependent on calcium ion. Couperwhite and McCallum (1974) observed that the addition of EDTA to batch culture media affected the ratio of D-guluronic acid to D-mannuronic acid in the alginate produced by Az_. vinelandii. Corpe (1964) observed increased polysaccharide, production by Chromobacterium violaceum in the presence of calcium. Calcium and Biofilms Calcium has been implicated in direct or indirect bridging between adjacent cell surfaces. Fletcher (1980) considers that calcium may act as a cross-linking or charge screening agent for the ionic groups in the EPS. Turakhia et al. (1983) were able to detach a mixed microbial film in a turbulent flow system with EGTA (a calcium-specific chelant) 14 and postulated that calcium was essential to the structural integrity of the biofilm. EDTA chelates divalent cations in the following preferential sequence: Ca2+ > Mg2+ > Sr2+ > Ba2+ However, disaggregation in EDTA might be connected with chelation of less strongly bound cations, as well as Ca . In recent years, however, EDTA has been superseded in many studies involving calcium (Cooksey and Cooksey, 1980; Turakhia et al., 1983) by EGTA, which can bind calcium over IO5 times (Reed and Bygrave, 1975) more effectively than it binds magnesium. It is likely that calcium plays a more important role in adsorp­ tion than any other cation. Calcium has a higher coordination number (7 or 8) and the coordination geometry is irregular in both bond angle and bond length. In both regards, calcium is quite different from magnesium, which maintains six coordination in a closely regular octahedron. Because magnesium requires a certain specific geometry, it is weakened in its ability to bind irregular geometries of coordination sites of biological molecules. Magnesium does not cross-link struc­ tures readily, for cross-linking usually demands a high coordination number and irregular geometries, which are characteristics of calcium. Biofilm Detachment At any point in the development of a biofilm under turbulent flow conditions, external portions of biofilm are sheared away into the fluid flow. L5 Detachment phenomenon can be arbitrarily categorized as erosion Or sloughing. Erosion refers to continuous removal of small portions of biofilm, which is highly dependent on fluid dynamic conditions. Under these circumstances, rate of detachment increases with increasing biofilm thickness and fluid shear stress at the biofilm-fluid interface (Trulear and Characklis, 1982). Sloughing refers to a random, massive removal of biofilm attributed to nutrient or oxygen depletion deep within the biofilm (Howell and Atkinson, 1976). Sloughing is more frequently witnessed with thicker, less' dense film which develops under low shear conditions. Detachment can also occur for reasons other .than hydrodynamic forces. Bakke (1983) has observed a massive detachment when substrate (lactate) loading to the biofilm was instantaneously doubled. Turakhia et al. (1983) and Characklis (1980) have observed increased detachment (of mixed microbial film) upon addition of chelants (EGTA and EDTA respectively) suggesting the importance of calcium to the cohesiveness of the biofilm. Many other chemicals (e.g., chlorine, bromine chloride, bromo-chloro-dimethyhydantoin, surfactants) have also been used for detachment with varying success. Most of the detachment work with chemicals has been monitored directly by measuring the changes in frictional resistance and/or heat transfer resistance. No significant work has quantified the amount of material remaining on the surface, identified the detached material, or quantified the amount of material detached. Detachment of biofilm is the major objective of many anti-fouling additives. Very little is known regarding the kinetics and the extent 16 of detachment. Such kinetic expressions would be useful for modelling purposes and as a comparative criterion for evaluating anti-fouling treatments. Organism Pseudomonas aeruginosa This organism has been studied extensively both in continuous flow stirred tank reactors (CFSTR), i.e., chemostats,-and in annular biofilm reactors. Information on growth rate and EPS formation rates are known under defined experimental conditions in chemostat and biofilm environ­ ments (Robinson et al., 1984; Bakke et al., 1984). Pseudomonas aeruginosa is a common waterborne polymer-forming bacteria capable of causing severe infections in a compromised host (Woods et al., 1980; Costerton, 1979). The primary mode of growth of Pis. aeruginosa in nature and disease is in polymer-enclosed microcolonies attached to a variety of surfaces. The polymer-enclosed, attached mode of growth purportedly protects Ps^ aeruginosa (and other biofilm organisms) from the bactericidal activity of bacteriophages and amoebae which are numerous in natural systems and from antibiotics and host defense mechanisms in diseased systems (Costerton, 1979). Psv aeruginosa can be considered a classic biofilm organism and for this reason is the bacterial species used in this study. Relevant characteristics describing Psv aeruginosa are as follows. a) gram stain: negative (Buchanan et al., 1974) b) morphology: rod shaped, typically 0.5 - 0.8 pm 17 c) metabolism: d) respiration: e) motility: f) polymer composition: by 1.5 - 3.0 pm (Buchanan et al., 1974) chemoorganotroph (Buchanan et al., 1974) strict aerobe (Buchanan et al., 1974) polar monotrichous flagellation (Buchanan et al., 1974) primarily mannuronic and glucuronic acids (Evans and Linker, 1973; Mian et al., 1978) 18 MATHEMATICAL DESCRIPTION OF THE SYSTEM Cellular production and polymer formation by Ps^ aeruginosa can be described mathematically using a mass balance approach. This section contains a mathematical model which describes biofilm processes, including accumulation and activity of bacteria immobilized in a biofilm and dispersed in the bulk phase. In this model the substrate carbon is partitioned into extracellular product and biomass carbon. For both of these processes, some substrate carbon is oxidized to carbon dioxide providing energy for synthesis of cells and EPS (Trulear, 1983). The annular reactor (Figure I) and the chemostat (Figure 2) were operated as continuous flow stirred tank reactors (CFSTR) in which bulk fluid concentration gradients do not exist. Accumulation of compounds in the bulk fluid can be described by a material balance of the general form: Accumulation of attached biofilm components can be described by a constitutive equation of the same general form as Equation I, but with net rate of accumulation net rate of + transport net rate of transformation (I) no transport term. 19 Removable slide Recycle Effluent Outer cylinder Figure I. A simplified diagram of the annular reactor. 20 SUBSTRATE INFLUENT TO ANNULAR REACTORS ANTI-BACKFLOW DEVICES MAGNETIC / STIRRING DISK Figure 2. A simplified diagram of the chemostat. 21 Chemostat Equation Cellular carbon. A mass balance across the chemostat for cellular carbon can be written as follows: V dx/dt net rate of cellular accumulation F (Xj, - x) net rate of cellular input by flow + jj x V rate of cellular production (2) where, V = volume of the system x = cellular carbon concentration t = time F = volumetric flow rate of the reactor, feed x. = influent cellular carbon concentration p = specific cellular growth rate (L3) (MxL-V 1). cellular carbon detachment rate from the biofilm Defining biofilm specific cellular growth rate, yfe, as: yb Xb. (6) where, specific cellular growth rate in the biofilm (t-1) The biofilm cellular carbon balance (Equation 5) can be written as: A dx /dt b Rdx A (7) BIOFILM WETTED SURFACE AREA N N ANNULAR REACTOR W Figure 3. The annular reactor (AR) was operated as a continuous flow stirred tank reactor (CFSTR). The insert describes the local environment in the reactor. The mathematical model was based or. this conceptual model. 24 Biofilm polymer carbon. A mass balance for the accumulation of polymer carbon can be written as follows: A dpb / dt A net rate of polymer carbon accumulation in the biofilm rate of polymer formation in the biofilm rate of polymer detachment from the biofilm (8) where, = polymer carbon density in the biofilm R = polymer carbon formation rate in the biofilm pb R = polymer carbon detachment rate from the biofilm (MpL 2) (M L-V 1) P (M L - V 1) P Polymer formation may be related to organism growth rate and population density according to the Luedeking and Piret equation for product formation (Luedeking and Piret, 1959; Robinson et al., 1984; Bakke et al., 1984). where, kb growth—associated polymer formation rate coefficient in the biofilm k'b nongrowth-associated polymer formation rate coefficient in the biofilm t y x " 1’ -I - L (MM t ) p x The biofilm polymer carbon balance (Equation 8) can be written: A Vdt substrate feed lines, and buffer feed lines were autoclaved. 6. The dilution water feed lines were connected to the pH controlling reservoir. Twenty-five milliliter of bleach was added into the cascade filters and the flow of dilution water was started. 7. The AR's were filled with 10% bleach in distilled water and operated in a batch mode for I hour. 8. The sterile capsule filter was connected to the AR and the flow of dilution water was started. 9. The substrate and buffer stock solutions were prepared and auto­ claved. 10. A minimum of 10-12 h (overnight) was allowed for residual chlorine to be flushed out from the annular reactors. 11. After 12 h, the substrate and buffer feed lines were connected to the AR's. 12. The experimental start-up procedure was initiated. Tubular reactor system. The TR system cleaning and sterilization was as follows: I. Following an experiment, twenty milliliters of bleach was added to the reactor and the TR was operated in the batch mode for 10 minutes. 44 2. The reactor was disassembled and cleaned with a soft bristle brush. 3. The reactor was assembled. 4. Before starting the experiment, stock solutions of glucose and buffer were prepared. 5. The dilution water and substrate feed lines were connected to the TR. 6. The TR was filled with the medium, and the flow rate in the recycle loop was adjusted. Experimental Start-Up BOD respirometer. The respirometer start-up procedure was as follows: 1. The sample bottles were inoculated with I ml of batch culture of Ps. aeruginosa through the septum on the bottles. 2. The electrolysis cell was filled with 2N to within 0.4 cm of the switch electrode. 3. The outer ground glass surface of the electrolysis cell was greased with silicone grease. 4. A sterile glass wool filter was placed in the stem of the electro­ lysis cell. 5. Six grams of KOH pellets were added to the alkali-container. 6. The electrolysis cell was placed on the alkali-container. 7. The current of the electrolysis cell was adjusted at 200 milliamps. 45 Chemostat. The experiment was started by inoculating the chemo- stat with I ml of batch culture of Pjs. aeruginosa. The chemostat was operated in a batch mode for 10-12 h . This usually resulted in a concentrated suspension of organisms in the 'chemostat. Substrate solution flow was then started at the desired dilution rate. The chemostat was operated in the continuous flow mode for at least six residence time prior to use in AR to allow the reactor to reach steady state. Annular reactors. Two parallel annular reactors were used for this study. The experiment was started by pumping organisms from a steady state chemostat culture of Ps_. aeruginosa (chemostat dilution rate 0.24 h *) into the AR’s for a period of 20-24 h (Experiments I & 2; 6 x IO^ cells ml *) and for 8-10 h (Experiments 3-7; 10^ cells ml )̂. This provided a defined microbial inoculum for the attachment experiments. Organism flow rates were maintained at 0.55 ± 0.05 ml min *. Substrate solution and dilution water were fed to the AR throughout the experiment. Additional buffering capacity was provided with Tris-HCl buffer (pH 7) during treatment with chelant to maintain pH at a constant value. EGTA or EDTA was added to the AR using step input and the concentration was kept constant for I hour. Two hours (twelve residence time) before chelant addition, the flow of calcium to the AR was stopped. Hence, the chelant demand in the liquid phase was negligible. Tubular reactor. The experiment was started by inoculating the TR with a mixed microbial population. The TR was allowed to operate in 46 the batch mode for 10-12 h. This usually resulted in a concentrated suspension of organisms in the IR. The recirculating pump was then started. The dilution water, substrate, and the buffer flow was also started. A biofilm was allowed to form over a period of several days before EGTA was added to the system. Additional buffering capacity was provided with Tris-HCl buffer (pH 7.5) during treatment with EGTA to maintain pH constant. Sampling a) glucose b) suspended solids Liquid samples. Liquid samples from . AR, TR, and the chemostat were directly collected from the effluent, prepared, and stored as follows: - filtered (Nuclepore Corp., Pleasanton, CA, No. 111107, average pore size 0.45 pm), samples were frozen until analysis - 25-100 ml samples were filtered (Nucle­ pore Corp., Pleasanton, CA, No. 111107, average pore size 0.45 pm), dried, and weighed - 4 ml samples were fixed in an direct equal volume of 4% sterile formalin, homogenized (Du Pont Co., Instrument Products Newton, CT, Sorvall Omni- .Mixer), and refrigerated until analysis by the method of Hobbie et al. (1977) c) acridine orange direct count (AODC) 47 d) carbon - filtered (Nuclepore Corp., Pleasanton, CA, No. 111107, average pore size 0.45 pm, 5 x 5 ml) and unfiltered samples (5 x 2 ml) were frozen until analysis. Biofilm samples. Biofilm samples were obtained from the AR removable slides. ■ Biofilm was scraped from the slide using a rubber policeman into 30-35 ml carbon-free distilled water. The resulting solution was homogenized (Du Pont Co., Instrument Products, Newton, CT, Sorvall Omni-Mixer) and subsamples were prepared and stored as follows: a) acridine orange — 2—4 ml samples were fixed in 4 ml of direct count sterile 4% formalin, and refrigerated (AODC) until analysis by the method of Robbie et al. (1977) b) carbon samples - 5 x I ml samples were frozen until analysis c) mass density - 15-25 ml samples were filtered (Nucle­ pore Corp., Pleasanton, CA, No. 11107, average pore size 0.45 pm), dried, and weighed. Analvtical Methods The hierarchy of analytical procedures performed on biofilm or reactor liquid are summarized in Figure 7. calculations CELL NUMBERS FILTRATION CELL CARBON x, xb SUBSTRATE CARBON, s p = TOC TOTAL ORGANIC CARBON BIOFILM OR REACTOR LIQUID SAMPLE Figure 7. Hierarchy of analytical procedures. 49 Suspended solids. Suspended solids concentration was determined by filtering 20-100 ml of reactor effluent through predried (103°C for I hour), preweighed, Nuclepore membranes (Nuclepore Corp., Pleasanton, CA, No. 11107, average pore size 0.45 pm). After filtration, the filter was dried at 103°C for I hour and weighed (Mettler Instruments Corp., Hightstown, NJ, Type H20 Digital Balance). Acridine orange direct count. Total number of cells in the effluent or the biofilm sample was determined by enumerating cells stained with acridine orange using epifluorescence microscopy (Leitz Wetzlar, Rochleigh, NJ, Ortholux II Universal Microscope) according to the method of Hobble et al. (1977). Carbon measurements. Carbon was used as the basis for the material balance relations which were used in this study. The carbon-containing components were as follows: glucose carbon, liquid phase cellular carbon, liquid phase polymer carbon, biofilm cellular carbon, and biofilm polymer carbon. Glucose carbon concentration. Glucose carbon concentration, s, was determined by measuring glucose concentration enzymatically using a modified version (Trulear, 1983) of Sigma 510 Glucose Analysis Procedure (Sigma Chemical Co., St. Louis, MO). The measured glucose concentration was multiplied by 0.4 g glucose carbon (g glucose) to determine glucose carbon concentration. Liquid phase cellular carbon concentration. The concentration of the liquid phase cellular carbon concentration, x, was calculated by 50 multiplying the acridine orange cell count and the average volume of __ ̂ the cell by the following quantities: 1.07 g cell cm (Doetsch et al., 1973; Bakken and Olson, 1983), 0.22 dry cell weight (wet cell weight) 1 (Luria, 1960; Bakken and Olson, 1983), and 0.5 g cell carbon (g cell dry weight) * (Doetsch et al., 1973; Luria, 1960). Liquid phase polymer concentration. The concentration of liquid phase polymer carbon, p, was calculated by measuring the total organic carbon (TOC) of a reactor liquid sample, TOCsoln> and performing the following calculation: p = TOC . - S - X ̂ soln TOC was determined using the ampule analysis module of the Oceanography International Carbon Analyzer (Oceanography International Corp., College Station, TX, Total Carbon System, Cat. No. 0524B). Biofilm cellular carbon density. The density of biofilm cellular carbon, x^, was determined by enumerating the total, number of cells per unit area using acridine orange cell count and performing calculations analogous to those presented for x. Biofilm polymer carbon density. The density of biofilm polymer carbon, p^, was determined by measuring the TOC of a known volume of biofilm suspension, TOC^, and performing the following calculation: Pb = T0Cb - Xb Total carbohydrate. Total carbohydrate on filtered and/or unfiltered effluent sample was determined using the phenol reaction 51 method of Hanson and Phillips (1981). These values were reported as I glucose equivalents. ' Total solids removed. Total amount of solids removed was deter­ mined by graphically integrating the curve describing the progression of effluent suspended solids during chelant addition. Dry biofilm mass. Dry biofilm mass was determined by filtering 15-25 ml of biofilm suspension through predried (103°C), preweighed, Nuclepore membranes (Nuclepore Corp., Pleasanton, CA, No. 11107, average pore size 0.45 pm). After filtration, the filter was dried at IOS0C for I hour and weighed (Mettler Instruments Corp., Hightstown, Nd, Type H20 Digital Balance). Total calcium. Total calcium in the influent and effluent samples was determined by EDTA titration. Total calcium in the dry biofilm sample was determined by the Chemistry Station-Analytical Lab (Montana State University) using atomic adsorption. Free calcium. Free calcium was determined with calcium activity electrode (Orion Research, Inc., Cambridge, MA., CA No. 93-20-00). C, H, and N analysis. Carbon, hydrogen, and nitrogen content of the lyophilized biofilm samples were determined using a Carlo-Erba elemental analyzer (Chemical Engineering Department, Montana State University, Bozeman). 52 Dissolved oxygen. Dissolved oxygen In the influent and effluent was measured using a YSI model-54 oxygen meter (Yellow Springs Instrument Co., Inc.). 53 RESULTS Comprehensive listings of raw data for all annular reactor, tubular reactor, and BOD respirometer experiments are listed in Appen­ dices A through E. The annular reactor was used for all studies involving Ps. aeruginosa biofilms. The tubular reactor was used only for mixed culture biofilms. The BOD respirometer was used for batch growth of Ps. aeruginosa. Cellular, polymer, and substrate data are reported as carbon equivalents. Substrate, glucose, was the sole carbon and energy source and the experiments were operated under substrate-limiting conditions. The annular reactor was operated at three different calcium concentrations; 0.4 (no added calcium), 25, and 50 mg I-1. Calcium concentration refers to total calcium and not free calcium, unless otherwise indicated. Influence of Calcium on Biofilm Formation The effect of calcium on cellular reproduction and polymer formation was measured by varying influent calcium concentration and measuring the changes in the following: I. biofilm cellular carbon areal density, Xfc ■2. biofilm polymer carbon areal density, p^ 3. biofilm mass 4. immobilized calcium 5. cellular carbon concentration, x 54 6. polymer carbon concentration, p 7. glucose carbon concentration, s 8. oxygen concentration, The effect of calcium on transformation or process rate was determined from the above measurements using the material balances described earlier. Table 7 presents a summary of experimental conditions for all AR experiments. Table 7. Summary of Experimental Conditions of Annular Reactor Exper­ iments . Expt # AR ' # D (h'1) Si (mg I *) Calcium (mg I"1) Total Free I 7 6 4.00 9.0 ND 8 6 3.84 9.0 ND 2 7 6 4.16 9.0 ND 8 - 6 3.84 . 9.0 ND 3 7 6 6.56 0.4 0.04 8 6 6.40 25.0 18.50 4 7 6 4.83 0.4 0.04 8 6 4.70 50.0 44.00 5 7 6 6.16 ' 25.0 18.50 8 6 6.00 . 25.0 18.50 6 7 6 6.32 50.0 .44.00 8 6 6.04 50.0 44.00 7 7 ■ 6 6.40 0.4 0.04 ND = not determined. 55 Effect of Calcium on Accumulation Rates Biofilm Mass The progression of biofilm mass for three different calcium concentrations in the dilution water is presented in Figure 8. The results indicate that the presence of calcium in the dilution water increases the rate of biofilm accumulation. However, there was no detectable difference in biofilm mass at 48 h for all calcium concen­ trations. The extent of biofilm accumulation at the end of six days as a function of calcium in the dilution wafer is presented in Figure 9. Accumulation was highest for experiments with 50 mg I * calcium. The results indicate that calcium influences the rate and extent of biofilm accumulation. The results suggest that calcium can be used for controlling biofilm accumulation. Biofilm Cellular Carbon The progression of biofilm cellular carbon for three different calcium concentrations in the dilution water is presented in Figure 10. The addition of calcium in the dilution water increases the accumula­ tion of biofilm cellular carbon. The extent of biofilm cellular accumulation (at the end of six days) increases with increasing calcium concentration in the dilution water (Figure 11). The results suggest that calcium can be used to attach more cells on a given surface. BI O FI LM M AS S (m g 56 3000 E2000 TIME (h) Figure 8. Progression of biofilm mass as a function of time for three different calcium concentrations. Lines drawn by observation. ( 0,O,$ 50 mg 1“1; B.O.# 25 mg 1~1; and 9,0,$ 0.4 mg 1~1) . 57 o IOOO CALCIUM IN DILUTION WATER (mg I"1) Figure 9. Change in biofilm mass (140 h) as a function of calcium in the dilution water. Error bars represent standard deviation of measurements of all experiments with same calcium concen­ tration in the dilution water. B IO F IL M C E LL U LA R C AR BO N A R E A L D E N S IT Y (m g rr r2 ) 58 O 80 160 TIME (h ) Figure 10. Progression of biofilm cellular carbon as a function of time for three different calcium concentrations. Curves represent time smoothed data. Parameters for time smoothed data were obtained using BMDP3R - nonlinear regression. ( 0,0,# 50 mg I-*; 8,0,0 25 mg I-1; and e ,O ,• 0.4 mg I-*). 59 Biofilm Polymer Carbon The amount of extracellular polymer accumulated within the reactor at the end of the experimental period (six days) was not influenced by calcium concentrations in the dilution water (Figure 11). There was no significant difference in biofilm polymer carbon (between experiments) at any time during biofilm development. Biofilm Calcium Calcium concentration in the biofilm increases with increasing calcium in the dilution water (Figure 12). Twelve residence times before sampling biofilm for calcium analysis, the flow of CaCO3 + HCL stock solution was stopped. Hence, there was no added calcium in the dilution water during the sampling period. Fixed and Volatile Biomass Fixed and volatile contents of the biofilm (140-146 h) at differ­ ent calcium concentrations are presented in Table 8. The data indicate that fixed and volatile biomass under different calcium concentrations remained the same. The fixed mass remaining after ignition generally represents the inorganic component of the biofilm and may include potassium, calcium, magnesium, sodium, iron, and trace amounts of cobalt, copper, molybdenum, and zinc. These elements are known to be required for metabolism. cellular carbon polymer carbon CM 25 50 CALCIUM IN DILUTION WATER (mg I"1) Figure 11. Change in biofilm cellular and polymer carbon (140 h) as a function of calcium in the dilution water. Error bars represent standard deviation of measurements of all experiments with same calcium concentration in the dilution water. 61 CALCIUM IN DILUTION WATER (mg I'1) Figure 12. Change in biofilm calcium as a function of calcium in the dilution water. 62 Table 8. Composition of Biofilm (Dry Weight Basis). Fixed Volatile Expt Calcium C H N Solids Solids tf - ■ . — Img I % % % % % 5 25.0 48.0 7.0 12.3 13.4 86.6 6 50.0 48.6 6.9 13.1 13.5 86.5 6* 50.0 47.8 6.9 12.8 12.8 87.2 7 0.4 47.5 6.9 12.4 13.6 86.4 * Analysis of biofilm left on the surface after EGTA addition. Biofilm Elemental Analysis Carbon, nitrogen, and hydrogen contents of the lyophilized biofilm samples under different calcium concentrations were determined using the Carlo-Erba elemental analyzer (Table 8). The data indicate that the overall biofilm elemental composition under different calcium concentrations remained the same. The results obtained are consistent with C, H, and N content of microbial cells reported in the literature, i.e., approximately 50% carbon, 20% oxygen, 10-15% nitrogen, and 8-10% hydrogen on a dry weight basis (Grady and Lim, 1980). Effluent Glucose Concentration Biofilm metabolic activity can be monitored indirectly by measur­ ing changes in effluent substrate (glucose) concentration. Figure 13 is a typical experimental progression of effluent glucose (carbon equivalents). At time zero, the effluent glucose concentration equals 63 C7» E UJ X O I-ZUJ ZD-I Ul U-UJ 0— 0— TIME (h) Figure 14. Change in effluent oxygen concentration as a function of time. Line drawn by observation (Experiment 6; AR//8). 66 Accumulation rate of substrate in the annular reactor was negligible (ds/dt = 0) during the sampling period (48-146 h). Incorporating this condition and substituting for and , Equation 15 can be written as: D(Si-S)V Xb A xb/s pb/s k\ pb/s (19) Specific substrate removal rate, qg, is defined as the left hand side of Equation 19: qs D (s. - s)V I Xb A (20) The specific substrate removal rate, when compared at a specific experimental run time, was highest for biofilm developed under 0.4 mg I-1 calcium in the dilution water (Figure 15). This difference in specific substrate removal rate was due to less accumulated biofilm cellular mass in the low (0.4 mg I *) calcium experiment. .As a result, more substrate was available per cell. There was no significant difference in specific substrate removal for biofilm developed under 25 and 50 mg I ̂calcium in the dilution water. Specific Oxygen Removal Rate Equation (18) accounts for oxygen utilization in the annular reactor. This equation was rearranged to calculate specific oxygen removal rate: (D+kc)(O2i-O2)V x, A b xb/o pb/o pb/o + (21) SP EC IF IC SU B ST R A TE R EM O VA L R A TE (g s ub st ra te (g c el l h) -1 ) 67 TIME (h) Figure 15. Change in specific substrate removal rate as a function of time with calcium in the dilution water as a parameter. Curves represent time smoothed data. (0.0,$ 50 mg 1-1; S 1D 1B 25 mg I-1; and 6,0 1B 0.4 mg 1-1). 68 The specific oxygen removal rate, q^, is defined as the left hand side of Equation 21: qo (D + kc)(02i - O2) V x, A b (22) Specific oxygen removal rate was highest for the biofilm developed under 0.4 mg I * calcium in the dilution water (Figure 16). Again, there was less attached cellular mass in the low calcium experiment which may account for higher specific oxygen removal rate. There was no significant difference in the specific oxygen removal rate for biofilm developed under 25 and 50 mg I ̂calcium in the dilution water. Specific Cellular Growth Rate in the Biofilm The specific cellular growth rate in the biofilm was calculated by dividing cellular carbon reproduction rate in the biofilm (R by biofilm cellular carbon concentration (x^). During the initial period of biofilm accumulation, the cells are growing near their maximum, growth rate (Figure 17). As reactor substrate concentration decreases (with increasing time), the specific cellular growth rate decreases as expected. Specific Cellular Detachment Rate The specific cellular detachment rate was calculated by dividing cellular carbon detachment rate (R^x) by biofilm cellular carbon (x^). The specific cellular detachment rate also decreases with time (Figure 18). At the end of the experimental periodj the specific cellular SP EC IF IC O XY G EN R EM O VA L R A TE (g o xy ge n (g c el l h) -1 ) 69 TIME (h) Figure 16. Change in specific oxygen removal rate as a function of time with calcium in the dilution water as a parameter. Curves represent time smoothed data. ( O ,• 50 mg I"!; D , ■ 25 mg I-1; and o 0.4 mg 1-1). B IO FI LM S P E C IF IC C E LL U LA R G R O W TH R A TE (I T 1) 70 TIME (h) Figure 17. Change in biofilm specific cellular growth rate as a function of time. Line drawn by observation. ( o ,• 50 mg 1-1; D 1B 25 mg 1“ 1; and o 0.4 mg 1~1 calcium). SP EC IF IC C E LL U LA R D ET A C H M EN T R A TE (h 71 TIME (h) Figure 18. Change in biofilm specific cellular detachment rate as a function of time. Line drawn by observation. ( o , • 50 mg 1-1; 25 mg I'1; and o 0.4 mg I"1 calcium). 72 detachment rate approaches specific cellular growth rate and the system approaches steady state. Specific Polymer Production Rate Specific polymer production rate was calculated by dividing polymer carbon formation rate (R^) by biofilm cellular carbon concentration (x, ). Specific polymer production rate decreases with D time because substrate concentration decreases (Figure 19). Specific Polymer Detachment Rate Specific polymer detachment rate was calculated by dividing polymer carbon detachment rate (Rdp) by biofilm polymer carbon concentration (p^). Specific polymer detachment rate decreases with time and appears to reach a steady state after 80 h (Figure 20). Effect of Chelant on Biofilm This section describes experiments wherein buffered EGTA and/or EDTA was used to detach Pis. aeruginosa and mixed culture biofilms. Results of detachment experiments with mixed culture biofilms were published in Applied and Environmental Microbiology, December 1983. Experimental conditions and summaries of the experimental results from AR and TR are given in Tables 9 and 10, respectively. Raw data from AR and TR experimental studies with the addition of chelant are presented in Appendices B and C. Twelve residences time (AR experiments) before the addition of EGTA, the flow of CaCO3 + HCl stock solution was stopped. Hence, there SP EC IF IC PO LY M ER PR O D U C TI O N R A TE (g p ol ym er (g c el l h) "1 ) 73 1.6 - TIME (h) Figure 19. Change in biofilm specific polymer production rate as a function of time. Line drawn by observation. (O,# 50 mg 1-1; 25 mg I-1; and O 0.4 mg I-1 calcium). SP EC IF IC P O LY M E R D ET A C H M EN T R A TE (h "1 ) 74 O 80 160 TIME (h) Figure 20. Change in biofilm specific polymer detachment rate as a function of time. Line drawn by observation. ( O ,• 50 mg 1-1; 25 mg 1-1; and O 0.4 mg 1-1 calcium). 75 Table 9. Summary of Annular Reactor Detachment Experiments. Expt # ' AR # Si mg I 1 Total calcium during biofilm accumulation* , -I mg I Chelant (EGTA) added mM Biofilm mass —2mg m Biomass detached —2mg m I 7 4.00 9.00 0.00 ND 0.0** 8 3.84 9.00I 0.25 ND 86.5 2 7 ,4.16 9.00 0.00 ND 0.0** 8 3.84 9.06 0.50 ND 260.0 3 7 6.56 0.40 1.00 919.5 116.3 8 6.40 25.00 1.00 1353.2 362.2 4 7 4.83 0.40 1.00(3 937.5 90.1 8 4.70 50.00 1.00 2772.53 313.0 6 7 6.32 50.00 1.00 2403.8 318.7 ND = not determined. @ Chelant was EDTA. * There was no added calcium in the dilution water during chelant addition. ** Control experiments; no additional removal of biomass from the surface. For experiments I & 2, the biofilm was allowed to accumulate for. four and five days, respectively, before chelant was added. For experiments 3, 4, & 6, chelant was added to the AR after six days of accumulation. Time of EGTA/EDTA addition was 60 min. 76 Table 10. Experimental Results. Conditions and Summary of Tubular Reactor Experiment #1 Experiment #2 Dilution rate (h *) 0.50 0.50 Influent (mg I ) substrate 17.60 17.60 Volume (I) 10.49 7.80 2Surface area (m ) 0.80 0.77 Tris (mM) 10.00 10.00 EGTA (mM) 1.00 1.00 Total biomass (mg) detached in one hour 325 865 Time of EGTA addition was 120 min. 77 was no added calcium in the dilution water during detachment studies. One hour (two hours in case of TR experiments) before the addition of EGTA or EDTA1 sufficient Tris-HCl was added to the system to give a final concentration of 10 mM Tris and the dilution water was also buffered with 10 mM Tris. Cessation of calcium feed and buffer addition did not perturb the system since no significant changes in the analytical characteristics of the effluent were detected (AR #7; Experiments I & 2). The experiment consists of adding a known concentration of EGTA or EDTA (step input) to the bioreactor and monitoring the changes in the following quantities in the effluent: suspended solids suspended cells glucose total carbohydrate Experimental progressions are presented in which time zero is defined as time at which EDTA or EGTA was added. Suspended Solids The effect of EGTA and/or EDTA on the biofilm was monitored by measuring changes in effluent suspended solids. The response of a biofouled annular reactor to the addition of a known concentration of EGTA (1.0 and 0.5 mM, respectively) is presented in Figures 21 and 22. The addition of EGTA resulted in the detachment of a portion of the biofilm as reflected by an increase in effluent suspended solids. Similar results were, obtained with all other AR experiments with Ps_. 78 EGTA # suspended solids O-------- -O detachment rate O - O - Q- - 140 70 O 20 40 60 TIME (m in) Figure 21. Response of a biofouled annular reactor (Experiment 6) to the addition of I mM EGTA. Lines drawn by observation. D E TA C H M E N T R A T E (m g m"2 m in"1 ) 79 suspended solids 0 - - - 0 detachment rote O i 3 0 20 IO O 3, TIME (min) Figure 22. Response of a biofouled annular reactor (Experiment 2; AR #8) to the addition of 0.5 mM EGTA. Lines drawn by observation. DETACHM ENT RATE (m g 80 aeruginosa biofilms. These results suggest that free calcium (or calcium-associated ligands of lower affinity than that of calcium for EGTA) is essential to the structural integrity of the Ps_. aeruginosa biofilms. EGTA was also added to the tubular reactor containing a mixed culture biofilm. The addition of EGTA resulted in the removal of mixed culture biofilm as reflected by the increase in suspended solids in the effluent (Figure 23). Similar results (Figure 24) were also obtained in a second experiment, where the biofilm mass was tenfold greater than in the first experiment. Thus, calcium is also essential to the structural integrity of mixed culture biofilms. A contiguous portion of the mixed culture biofilm from the tubular reactor was scraped from the surface and suspended in the growth media. Disaggregation of the biofloc was observed under a phase contrast microscope when EGTA was added. Detachment Rate A material balance approach was used to determine biofilm detachment' rate from the surface. A mass balance (for suspended solids) across the reactor is as follows: V dx,T dt - F x,T + rD A (23) net rate of accumulation in the reactor net rate of output by flow net rate of detachment where, V = volume of the system S U S P E N D E D SO LI D S (m g I 81 Suspended Solids Detachment Rate T I M E ( m i n ) 3 0 20 IO 0 Figure 23. Response of a biofouled tubular reactor (Experiment //I) to the addition of I mM EGTA. Lines drawn by observation. D E TA C H M E N T R A TE (m g m ~2m in~i) S U S P E N D E D S O LI D S (m g I" ') 82 S u s p e n d e d S o l i d s D e t a c h m e n t R a t e — O IOO 50 Figure 24. Response of a biofouled tubular reactor (Experiment #2) to the addition of I mM EGTA. Lines drawn by observation. D E TA C H M E N T R A TE (m g m -2m in~i) 83 A = surface area (L2) , -3V XT = suspended biomass concentration (ML ) t = time (t) rD = net detachment rate ( M L ' V F = influent flow rate (lV 1) The detachment rate can be calculated by rearranging Equation 23 as follows: Rd = (V dx^/dt = F xT) / A (24) The detachment rate, Rq, in these experiments (AR's and TR's) was calculated using Equation 24. The addition of chelant resulted in the removal of Ps. aeruginosa and mixed culture biofilm in the first 5-10 minutes (Figures 21-24). Suspended Cells and Polymers The suspended biomass resulting from detachment was composed of cellular and polymer biomass. The addition of EGTA to annular reactors containing Pis. aeruginosa biofilm resulted in partial removal of cellular and polymer components of the biofilm as measured by the increase in their respective constituents (Figures 25 and 26) in the effluent. Total carbohydrate was used to estimate the extracellular polymer content in the effluent. Similar results were obtained with all Ps. aeruginosa biofilms (AR experiments) and mixed culture biofilms (TR experiments)-. SU SP EN DE D C EL LS /m l 84 TIME (min) Figure 25. Change in suspended cell concentration in the effluent of the annular reactor (Experiment 2; AR #8) after EGTA (I mM) addition. Line drawn by observation. Error bars represent standard deviation of measure­ ments of the same sample. 85 LUI-< OCQ >- X O OQ OC< O g OI- 0 20 4 0 60 T I M E (min) Figure 26. Change in total carbohydrate (as glucose equivalents) in the effluent of the annular reactor (Experiment 2; AR #8) as a result of EGTA (I mM) addition. Line drawn by observation. 86 The calculated suspended cell mass in the effluent was consistent­ ly much smaller than the suspended solids mass (Figure 27), which suggests that the detached biofilm mass was comprised of material other than bacterial cells. This is consistent with the findings of Trulear (1983) who showed that biofilm is largely extracellular polymer sub­ stance (EPS). The results of these and other (Costerton et al., 1978) experiments show that EPS is largely carbohydrate in nature (Figures 26 and 28). Microbial Activity The effect of chelant (EGTA or EDTA) on microbial activity was monitored by measuring the changes in effluent glucose concentration, an in situ measure of microbial activity. There was no significant change in effluent glucose concentration as a result of chelant (EGTA and EDTA) addition (AR experiments) indicating that chelant did not affect overall microbial substrate utilization (Table 11). Thus, microbial activity was presumably unaffected. There was also no change in the effluent glucose concentration as a result of EGTA addition to mixed culture biofilms (Appendix B). Total Biofilm Detached The total amount of biomass detached as a result of chelant addition was determined by integrating the curve (e.g., Figures 21-24) describing the progression of effluent suspended solids. The amount of biofilm detached in the various experiments are reported in Tables 9 and 10. There was no significant difference in the amount of biofilm Table 11. Change in Effluent Glucose Carbon Concentration as a Result of Chelant Addition (AR Experiment 4). Time after chelant addition (min) Effluent glucose carbon AR #7* (mg I S AR #8** 0 0.58 0.64 15 0.70 — 20 — 0.64 30 0.80 — 40 0.80 — 50 — 0.67 60 0.80 0.67 AEDTA (1.0 mM). AAEGTA (1.0 mM) . 88 Time (min) Figure 27. Change in suspended solids and suspended cell mass in the effluent due to the addition of I mM EGTA. Lines drawn by observation. Error bars based on standard deviation from epifluorescent cell counts. 89 TIME (min) Figure 28. Change in total carbohydrate (as glucose equivalents) in the effluent of tubular reactor (Experiment //I) after the addition of I mM EGTA1 Line drawn by observation. 90 removed when either EGTA or EDTA was added to the Psv aeruginosa biofilm developed in low calcium (0.4 mg I conditions. The amount of mixed culture biofilm removed in the second experiment was significantly higher than the first experiment largely because the biofilm in the second experiment was allowed to form over a much longer period. Biofilm Calcium The change in biofilm calcium concentration before and after the addition of EGTA and/or EDTA (Table 12) indicates a significant decrease in calcium concentration in the Psv aeruginosa biofilm occurs as a result of chelant addition. Table 12. Biofilm Calcium Before and After Chelant (EGTA) Addition. Calcium in Biofilm calcium (pg g *) Expt AR dilution water i Before chelant After chelant # if n -L mg I addition addition 3 I 0.4 ND 55 3 8 25.0 1250 H O 4 7 0.4* 388 117 4 8 50.0 11200 800 ND = not detectable. * Chelant was EDTA. 91 Batch Growth Experiments . The effect of calcium on maximum specific growth rate of Ps. aeruginosa was determined by measuring the oxygen uptake in batch (suspended) cultures using BOD respirometer. The BOD respirometer monitors oxygen uptake in three different batch cultures simultaneous­ ly. Two different calcium concentrations, 3.2 (no added calcium) and 25 mg I , were tested. Raw data from batch growth experiments are presented in Appendix E. Experimental conditions and summaries of experimental results are given in Table 13. Table 13. Experimental Conditions and Summary of Batch Growth Experi­ mental Results. Expt # Unit # Si mg I 1 Total calcium ..-I mg I Free calcium . — I mg I "m T 1 Ys/o SC(SO2)-1 I I 80 25.0 . ND 0.386±0.004 0.80 . 2 80 3.2 ND 0.318+0.003 1.00 3 80 3.2 ND 0.334+0.004 0.89 2 I 400 . 25.0 12.50 0.432±0.003 ND 2 400 3.2 0.02 0.32510.001 ND 3 400 3.2 0.02 0.32910.001 ND 3 I 400 25.0 12.50 0.349+0.002 0.89 2 400 25.0 12.50 0.36410.003 0.89 3 400 3.2 0.02 0.308+0.002 0.89 ND = not determined. The experiments were initiated by inoculating the sample bottles with a batch culture of .Ps. aeruginosa. Calcium was added to the sample bottles using sterile disposable syringes from a sterile CaCO^ + 92 HCl stock solution. The sample bottles were continuously stirred to facilitate the transfer of oxygen to the liquid. The temperature of the system was maintained at 24°C. Microbial growth and reproduction was monitored by measuring cumulative oxygen uptake. The progression of oxygen uptake (Figures 29-31) is frequently referred to as "bacterial growth curve." The cumulative oxygen upta