Biofilms on stainless steels exposed to process...

1Introduction

Biofilms on stainless steels

exposed to process waters

Katri Mattila

Department of Applied Chemistry and Microbiology

Division of Microbiology

University of Helsinki

Finland

Academic Dissertation in Microbiology

To be presented, with the permission of the Faculty of Agriculture and Forestry of the

University of Helsinki, for public criticism in auditorium 2041 at the University of

Helsinki Biocenter, Viikinkaari 5, on August 26,

at 12 o´clock noon.

Helsinki 2002

2 Introduction

ISSN 1239-9469ISBN 952-10-0309-X

ISBN 952-10-0310-3 (pdf)

Electronic publication available at http://ethesis.helsinki.fi

YliopistopainoHelsinki, Finland 2002

Front cover: Hybridized Baltic Sea biofilm seen by CLSM (Fig 14b) surrounded bypaper mill biofilm forming bacteria seen by SEM (Fig 9).

3Introduction

To my family

Supervisor: Professor Mirja Salkinoja-SalonenDepartment of Applied Chemistry and MicrobiologyUniversity of Helsinki, Finland

Reviewers: Docent Anja Klarin-HenricsonElektrowatt-Ekono, Finland

Docent Kirsten JørgensenResearch Programme for Environmental TechnologyFinnish Environment Institute, Finland

Opponent: Professor Wolfgang E. KrumbeinUniversität Oldenburg, Germany

ContentsList of original papers ....................................................................................... 7The author’s contribution ................................................................................. 7Abbreviations ................................................................................................... 81 Introduction........................................................................91.1 Biofilms ..................................................................................................... 91.1.1 Heat exchangers and paper machines as biofouling targets .................... 9Heat exchangers as an environment for microbial growth ................................. 9Paper machines as an environment for microbial growth ................................. 101.1.2 Biofilm: growth on surfaces .................................................................. 11The role of quorum sensing in biofilm formation ............................................. 13Factors influencing the attachment of bacteria onto solid substrata .................. 14Mature biofilm and detachment ...................................................................... 151.1.3 Cultivation of biofilms .......................................................................... 151.1.4 Methods for biofilm characterization ................................................... 16The confocal laser scanning microscope ......................................................... 16Study of biofilm microbial communities by in situ hybridization ..................... 18Scanning electron microscopy as a tool to study biofilms ................................ 20Energy dispersive X-ray analysis (EDS) for analysing elemental composition .... 212 Biofilm-driven processes ............................................................................. 222.1 Biofouling ................................................................................................ 22Measurement of the biomass attached to surfaces ........................................... 222.2 Biocorrosion ............................................................................................ 24Methods for the detection of (bio)corrosion ..................................................... 252 Aims..................................................................................273 Materials and methods .....................................................283.1 Stainless steel types ................................................................................. 283.2 Experimental set-ups used in this biofilm study ....................................... 283.3 In situ hybridization ................................................................................ 31....................................................................................................................... 314 Results and discussion ......................................................334.1 Properties of biofilms as displayed by scanning electron microscopy, SEM 334.1.1 Visualising the formation of biofilms on stainless steel in the Baltic Sea

water and in the paper machine wet end ................................................... 334.1.2 Biofouling of stainless steel surfaces exposed in the Baltic Sea and in a

paper machine ........................................................................................... 35Conclusions on biofilm formation as observed by SEM ................................... 354.1.3 Corrosion- related microbiologial events on stainless steel in Baltic Sea

water: factors influencing ennoblement ..................................................... 354.1.4 Biological events linked to the corrosion of stainless steel in a paper

machine ..................................................................................................... 41

6 Introduction

4.2 Properties of biofilms on stainless steels as displayed by confocal laserscanning microscopy (CLSM) ..................................................................... 46

4.2.1 In situ hybridization in a study of biofilm forming bacteria .................. 464.2.2 Use of autofuorescence and specific fluorescent stains for analysing the

biofilms formed on stainless steel .............................................................. 474.2.3 Properties of the biofilm clusters grown on stainless steel .................... 47Conclusions about biofilm formation observed by CLSM ................................ 524.3 Methods developed in this study.............................................................. 525 Summary and conclusions ................................................586 Tiivistelmä .........................................................................597 Acknowledgements ...........................................................608 References ........................................................................61

7Introduction

List of original papers

Mattila K, L Carpén, T Hakkarainen and MS Salkinoja-Salonen. 1997.Biofilm development during ennoblement of stainless steel in Baltic Seawater: a microscopic study. International Biodeterioration andBiodegradation, 40: 1-10.

Mattila K, L Carpén, L Raaska, H-L Alakomi, T Hakkarainen and MSSalkinoja-Salonen. 2000. Impact of biological factors on the ennoblementof stainless steel in Baltic Sea water. Journal of Industrial Microbiologyand Biotechnology, 24: 410-420.

Mattila K, A Weber and MS Salkinoja-Salonen. 2002. Structure and on-site formation of biofilms in paper machine water flow. Journal of IndustrialMicrobiology and Biotechnology, 28:268-279.

Uutela P, K Mattila, L Carpén, L Raaska, T Hakkarainen and MS Salkinoja-Salonen. 2002. Biogenic thiosulfate and oxalate in paper machine depositsconnected to corrosion of stainless steels. International Biodeteriorationand Biodegradation, in the press.

The original papers have been reproduced with kind permission of the copyright holdersNature Publishing Group and Elsevier Science.

The author’s contribution

Paper 1:Katri Mattila carried out all the experimental work except for the opencircuit measurements. She interpreted the results, wrote the paper, andprepared all the figures.

Paper 2:Katri Mattila wrote the paper and is the corresponding author of the paper.She performed all the experimental work except for the open circuitmeasurements, viable microbial counts, CTC-DAPI stainings, and a partof the acridine orange stainings. She prepared all the figures.

Paper 3:Katri Mattila wrote the paper and is the corresponding author of the paper.She carried out all the experimental work and prepared all the figures.

Paper 4:Katri Mattila carried out the epifluorescence microscopy, scanning electronmicroscopy and elemental analysis. She developed the microbiological partof the experimental set-up, and was responsible for that part of theinterpretation.

II

IV

I

III

8 Introduction

AAAAA

BBBBB

hhhhh

ccccc

sssss

Abbreviations

AFM atomic force microscopeATP adenosine triphosphateCBE Center for Biofilm Engineering, Montana State UniversityCIP cleaning in placeCLSM confocal laser scanning microscopyCMC carboxymethyl celluloseCon A concanavalin A, in this study labelled with tetramethylrhodamineCTC 5-cyano-2,3-ditolyl tetrazolium chlorideDAPI 4´,6-diamidino-2-phenylindolDNA deoxyribonucleic acidDSMZ Deutsche Sammlung von Microorganismen und Zellkulturen GmbHEDS energy-dispersive X-rayEDTA ethylenediamino tetraacetic acidEPS exopolysaccharideEtBr ethidium bromideFESEM field emission scanning electron microscopeFISH fluorescent in situ hybridizationGFP green fluorescent proteinHSL homoserine lactonelog K

owlogarithm of the n-octanol: water partition coefficient

LPS lipopolysaccharideMIC microbially influenced corrosionNA numerical apertureOMP outer membrane proteinsRNA ribonucleic acidSDS sodium dodecyl sulphateSEM scanning electron microscopySPM scanning probe microscopeSRB sulphate-reducing bacteriarfu relative fluorescence unitTEM transmission electron microscopyUNS unified numbering system

9Introduction

AAAAA

BBBBB

1 Introduction

1.1 Biofilms

During the millions of years when bacteriawere the only form of life on Earth, theprevailing aquatic environment wasextremely oligotrophic. The nichespermissive for life were limited by hostileenvironmental factors like UV radiation,heat and acidity (Costerton and Lappin-Scott 1995). At that time the purpose of theplanktonic, free-living mode of bacterialgrowth was to enable them to move fromone habitat to another until a nichepermissive for growth was found. Biofilmformation allowed these first sessileorganisms to remain in place, and to trapand utilize the scarce organic compounds.The development of co-operation amongbacterial groups permitted the use of morecomplex or more refractory nutrients.Biofilm formation also changed theenvironment of the colonized surface andmade it more suitable for the bacteriagrowing there. Biofilm formation hastherefore been a means of survival forbacteria, and is the reason why bacterialbiofilms are characteristic of the life formsfound in extreme environments like stonesurfaces in hot springs or the surface ofstones in cold, oligotrophic mountainstreams. Nowadays this survival strategyis successfully used by bacteria in industrialsystems such as drinking water distributionnetworks, (Szewzyk et al 2000,LeChevallier 1999, Koskinen et al 2000),paper machines (Väisänen et al 1994, 1998,Claus and Müller 1996), dairies (Pirttijär-vi et al 1998), breweries (Storgårds 2000)and power plants (Linhardt 1996). Intechnical systems, biofilm formation cancause problems like the contamination ofdrinking or process water by micro-organisms, an increase in the flowresistance of pipe lines or a decrease in thethermal transfer capacity in heat exchangers

(Flemming 1996). The presence ofpathogenic bacteria in biofilms in drinkingwater distribution systems is a health risk(Legionella pneumoniae, Klebsiellapneumoniae, Mycobacterium avium,LeChevallier 1991 and 1999, Szewzyk etal 2000).Despite the frequent problems caused bybiofilms, there are industrial systems wherebiofilm growth is a beneficial phenomenon.The proper functioning of activated sludgeplants, trickling filters or anaerobicdigesters is based on functional biofilms.In this thesis the main focus is onundesirable process biofilms.

1.1.1 Heat exchangers and papermachines as biofouling targets

Biofouling is a general term that refers tothe undesirable accumulation of bioticdeposits on a surface. Microbially induceddeterioration occurs in many industrialprocesses. Damage to metal materials isknown to occur in aqueous media, wherespecific microorganisms can findconditions favourable for growth(temperature, nutrients) (Weber and Knopf1996). This often happens in heatexchangers and paper machines. Both ofthese processes use large amounts ofnatural water, are kept at a constanttemperature, and offer a range of substratesfor microorganisms to adhere to and growon.

Heat exchangers as an environmentfor microbial growth

Industrial coolers are complicated systemsof pipes in which excess heat is transferredthrough metal surfaces to colder, circulatingnatural water. Cleaning in place (CIP) with

10 Introduction

Pick-up felt

Splash area of wet end

Spray water

chemicals is the main method used forremoving unwanted biofilms. Heatexchangers are usually made of stainlesssteel to minimize corrosion.

Paper machines as an environment formicrobial growth

Paper is made of wood fibre. In addition,paper manufacturing uses water, chemicaladditives such as fillers, binders, mineralpigments, CMC or starch, and resins orneutral adhesives for sizing. Chipped woodis broken up into wood fibres mechanically(ground wood, mechanical pulp) orchemically (chemical pulp). The pulp isslurried with additives and water to form a“stock” for papermaking (Fig 1). The stockin the head box of a paper machine contains0.3-0.6 % w/v of fibre and more than 99%water (Biermann 1996). From the head boxthe stock is spread onto the wire to form aformatted paper web. Water is removedfrom the web first by gravity, then bysuction (consistency 18-23 %) and by

pressure (35-55 %), and finally by heat(more than 90 % d.wt, Biermann 1996).Different types of paper machine have beendesigned to produce different grades ofpaper. Modern paper machines can be upto 600 m long, and the web up to 14 m wide.The paper web formed on the wire movesat a speed of up to 1500 m min-1 (≈ 90 kmh-1). The annual production capacity of onepaper or board machine in Finland isaround 300 000 metric tons.The presence of water and oxygen in thepaper machine, the elevated temperature(30-55ºC), and suitable pH conditions (4-10), permit the growth of many kinds ofmicroorganism (Väisänen et al 1994 and1998, Lindberg et al 2001a, Kolari et al2001, Busse et al 2002). Bacteria, mouldsand yeasts may deteriorate the rawmaterials used in paper making (Väisänenet al 1998, Pirttijärvi 2000 and 2001,Salzburger 1996), form biofilms which clogscreens, wires and felts, or generatemetabolic products that are corrosive orotherwise deteriorate the machinery.

Figure 1. Basics of a paper machine. Paper making stock is pumped into a head box, from whereit is spread onto a moving wire (=finely woven plastic web). Water drains through the wire bygravity and suction. The fibres remain on the wire to form a paper web. Fibres tend to alignthemselves in the direction of movement of the wire (=machine direction). The paper web at thisstage contains 80 to 90 % of water. The wire returning to the head box is washed with a pressurizedspray water. The paper web is pressed between dryer felts and up to press roll nips. Between thetwo pressuring rolls water is pressed from the web to reach a dry matter content of ca. 40 % w/w.Pressing improves inter-fibre bonding by bringing the fibres closer together. The paper web thenpasses between several steam-heated drying cylinders until the dry weight of ≥ 90 % is reached.This drying is the most energy consuming part of papermaking. After drying the paper is woundonto spools to form machine rolls (modified from Papermaking, FINNPAP lecture material, Hel-sinki, Finland, 1995).

11Introduction

The environment inside the paper machinesand heat exchangers may promote thegrowth of certain microbial species atspecific locations, even though the overallconditions in the installations do not favoursuch microorganisms (Weber and Knopf1996). Biofilm formation is one of the mostcommon causes for a local change in theenvironment inside industrial processesusing water.

1.1.2 Biofilm: growth on surfaces

Non-living surfaces in an aqueousenvironment rapidly accumulate organicmolecules and inorganic ions to form a layercalled the ”conditioning film” (Fig 2).

Therefore, in all but the most oligotrophicecosystems, planktonic bacteria actuallyadhere to the surface of a surface film thatmay have different chemical propertiesthan the non-living surface (Costerton andLappin-Scott 1995, Marshall 1997). Theinitial attachment of bacteria to a surfacefrequently involves a portion of the cell, aflagellum or EPS, while the bacterial cellcontinues to revolve (Marshall 1988).During so-called reversible attachment, thebacteria use its motility to sustain contactwith the surface while searching for asuitable location there (Korber et al 1995).Such a search is also called chemosensoringif the bacteria prefer specific substratespresent on the surface or produced by other

Figure 2. A) The initiation of biofilm growth. At stage 1 a conditioning film accumulates on asubmerged surface. Later, at stage 2, planktonic bacteria from the bulk water colonize the surfaceand begin a sessile existence by excreting exopolymers anchoring the cells to the surface (modifiedfrom Geesey 1993).

Figure 2. B) Conceptual model of a mature bacterial biofilm drawn on the basis of CLSMexaminations of a large number of mono- and mixed-species biofilms. The discrete microcoloniesof microorganisms are surrounded by a network of interstitial voids filled with water. The arrowsindicate convective flow in the water channels (Costerton et al 1994, Courtesy of Center forBiofilm Engineering, Montana State University, Bozeman).

12 Introduction

bacteria (Nielsen et al 2000). Methods ofsurface translocation like the flagellarmotility of bacteria (Davey and O’Toole2000, Martínez et al 1999, McBride 2001),are called positioning mechanisms.

Flagellum-mediated motility is required inEscherichia coli for both approaching andmoving across the surface. Type I pili andouter membrane proteins are required fora stable organism-surface interaction in E.coli (Fig 3). Pseudomonas aeruginosa usesits flagellum only for bringing the cell intothe proximity of a surface (Davey andO’Toole 2000). The initially reversibleattachment may transform into anirreversible one, or to detachment of the

bacterial cell. Reversible attachment isconsidered to be predominant in nature(Korber et al 1995).

The initial attachment of a single bacteriumis based on interaction between the cell andthe substratum. The subsequent growth andmaturation of the biofilm depends on cell-to-cell interactions called coaggregation.Coaggregation can be defined as “therecognition and adhesion betweengenetically distinct bacteria” (Whittaker etal 1996). The formation of a biofilm cluster,coaggregation, may be regulated by meansof signalling systems between cells sharingthe same location, irrespective of whetherthey are related or not (Xie et al 2000).

Figure 3. Development of biofilm by gram-negative bacteria. This figure outlines the currentmodel for the early stages in biofilm formation in P. aeruginosa and E. coli.A) In P. aeruginosa, flagella are required to bring the bacterium into the proximity of a surface.LPS mediates early interactions, with an additional possible role of outer membrane proteins(OMPs). Once the bacteria are in a monolayer on the surface, type IV pilus-mediated twitchingmotility is required for the cells to aggregate into microcolonies. The documented changes ingene expression at this early stage include the upregulation of alginate biosynthesis genes anddownregulation of flagellar synthesis. Production of cell-to-cell signalling molecules (acyl-HSLs)is required for the formation of a mature biofilm. Alginate may also play a structural role in thisprocess.B) In E.coli, flagellum mediated swimming is required for both approaching and moving acrossthe surface. Organisms-surface interactions require type I pili and the outer membrane proteinAg43. Finally, EPS known as colanic acid is required for development of the normal E. colibiofilm architecture (modified from Davey et al 2000).

13Introduction

lasI mutant wild-type

The role of quorum sensing in biofilmformation

Quorum sensing means cell-to-cellsignalling, and is an important phenomenonof microbial communities. Recent results(Davies et al 1998, Whitehead et al 2001)show that it regulates biofilm formation.Two quorum sensing systems have beendescribed for Pseudomonas aeruginosa(Davies et al 1998, Wimpenny et al 2000):lasR-lasI, which controls virulence but alsoregulates the expression of a second system,and rhlR-rhlI, which is involved in theformation of secondary metabolites. Thesesystems control the production of the signalmolecules: butyryl homoserine lactone byrhlI, and 3-oxododecanoyl-homoserinelactone by lasI.The wild type and a lasI mutant bothattached themselves to a surface, whereasthe lasI mutant retained its planktonicbehaviour by forming a thin, evenlydistributed biofilm that was dispersed in 5min by detergent treatment (SDS, 0.2%)(Davies et al 1998, Fig 4). Wild type cellsof P. aeruginosa formed characteristicmicrocolonies composed of groups of cellsseparated by water channels. Similarmicrocolonies were formed by the lasI

mutant, but only after addition of anautoinducer (3-oxododecanoyl-homoserinelactone).Quorum sensing requires a sufficientdensity of bacteria. For this reason, neitherof the P. aeruginosa signal moleculeswould be expected to participate in theinitial stages of biofilm formation, i.e.attachment or proliferation. Instead, thesequorum sensing signals may be involvedin biofilm differentiation. The resultssuggest that the initial stages of biofilmformation in the lasI mutant, proceed in asimilar fashion to that of the wild type, butthat differentiation from attachedplanktonic cells into biofilm bacteria wasblocked.This field has been intensively investigatedever since the first regulatory signallingsystem was found (Shapiro 1998,Kolenbrander 2000, Miller and Bassler2001, Whitehead et al 2001, Schauder andBassler 2001). Quorum sensing systemshave been found to play an important rolein mature biofilm communities. Signallingsystems regulate the morphologicaltransformation of bacteria when theybecome a permanent part of the biofilm:flagellin synthesis is decreased, and theproduction of exopolysaccharide is

Figure 4. Characteristics of P.aeruginosa wild type and quorum-sensing mutant biofilms. lasI mutantforms an evenly distributed biofilm,which was detached and dispersed in5 min following treatment with sodiumdodecyl sulfate (SDS, 0.2%) (Courtesyof Center for Biofilm Engineering,Montana State University, Bozeman).

14 Introduction

increased (Watnick and Kolter, 2000).Polysaccharide production is favouredsince it reinforces the biofilm structure(Danese et al 2000). Cells located in thecentre of a biofilm cluster do not divide atall, or divide only slowly. However, theyremain viable and culturable once freedfrom the plastic encasement of the biofilmcluster (Geesey 2001).

Bacterial products diffusing from one cellto another cell may be responsible forintercellular communication. The reasonwhy metabolic communication, likequorum sensing, is not efficient in the caseof planktonic bacteria, is that the diffusingmolecules will be diluted in the aqueousphase and only a small amount may reachthe neighbouring bacteria (Davey andO´Toole, 2000).

Factors influencing the attachment ofbacteria onto solid substrata

The major factors regulating attachment ordetachment of bacteria onto or fromsurfaces are nutrient availability, theelectrochemical properties of the surface,and liquid flow. When nutrients are non-limiting in the liquid phase there is no needfor the bacteria to attach themselves. Stresssituations like a depletion of nutrientsmakes sessile growth more favourable inflowing liquids (O’Toole et al 2000).Adding biocides at sublethal concentrationsinto process waters may be a stress factorthat initiates biofilm formation (M. Kolari,unpublished data), even though the desiredeffect was the opposite. Adherence to asubstratum is favoured in cases where thesubstratum could be used as a source fornutrients (Watnick et al 1999). ”Over”growth of a biofilm cluster may causedetachment. Free-living bacteria will searchfor a favourable environment in which toreattach themselves.The attachment to non-living surfacesdepends on the physicochemical properties

of both the bacteria and the substratum(Dalton et al 1994). Non-living surfaces canbe divided into two main classes based ontheir surface energy: 1) High-energysurfaces, which are usually also hydrophilicand frequently negatively charged. Theseinclude inorganic materials such as glass,non-noble metals or minerals. 2) Low-energy surfaces are often hydrophobic andhave a low negative or positive electrostaticcharge. These include synthetic polymers,noble metals and high quality stainlesssteels. Because of the high surface activity,high-energy substrata readily adsorbdissolved solutes or atmosphericcontaminants and are therefore rarely clean(Fletcher 1990). Contaminants that form aconditioning layer change the properties ofthe naked substratum surface and thereforealso the surface charge. Bacteria maychange their surface composition inresponse to the environment. The degreeof surface hydrophobicity of the microbialcell has been used to predict attachmentphenomena. However, its relativeimportance is likely to be low because 1)no clear trend has been found between cellattachment and the measured surfacehydrophobicity, and 2) cell surfacehydrophobicity is not necessarily constantfor a given organism (Scheuerman et al1998).

The topography of the substratum and theliquid flow influence the attachment ofbacteria. It is assumed that the smootherthe surface topography and the higher theliquid flow, the more difficult it is formicrobes to become attached. Scheuermanet al (1998) reported that the presence ofgrooves in a surface, running perpendicularto the direction of the flow, resulted inpreferential attachment of bacteria to thedownstream edges of the grooves, and to alesser extent to the upstream edges. Non-motile bacteria were not found in thebottom of the grooves. An increase in flowrate (from 28 mm s-1 to 83 mm s-1) promotedattachment and decreased the importance

15Introduction

of the positioning of the cells relative tothe grooves. The flow rate recommendedby Geesey (1993) for the prevention ofbacterial attachment and subsequent risk ofcorrosion is > 1.5 m s-1. According toStoodley et al (1999), a mixed-speciesbiofilm grown in a laminar flow consistedof roughly circular-shaped microcoloniesseparated by water channels. In contrast,biofilm microcolonies grown in a turbulentflow were elongated in the downstreamdirection, forming filamentous “streamers”.These extensive, ripple-like structuresallowed the biofilm to migrate downstream.

Mature biofilm and detachment

Biofilms that have reached a stable size(thickness), i.e. where growth iscompensated by detachment, are calledmature. Mature biofilms can have acomplex architecture. Early studies onbiofilms by electron microscopy requireddehydration of the specimens, leading to adeceivingly simplistic view of biofilms ascells piled on top of one another (Costertonet al 1995). Recent advances in confocalscanning laser microscopy have allowedvisual inspection of fully hydrated biofilms.Visual observations of the biofilm interiorhave been connected to chemical data withmicroelectrode studies (Rodrigues et al1992, DeBeer et al 1997, Costerton et al1994 and Santergoeds et al 1999). The useof a microelectrode with a small diametertip (down to 10 mm) enabled separation ofchemical data obtained either from thebiofilm clusters or the void space seen byCLSM. This has radically changed ourviews of biofilm architecture (Lawrence etal 1991, O’Toole et al 2000). Maturebiofilms are highly hydrated openstructures containing a high proportion ofexopolymers and large void spaces betweenthe microcolonies (Fig 2b) (Lawrence et al1991). The void space will remainuncolonized if there is a low substrateconcentration or/and a high signal moleculeconcentration. Similarly, a high concentration

of leaking metabolites will direct the free-swimming bacteria towards themicrocolonies (Tolker-Nielsen et al 2000b).Stewart (1993) divided biofilm detachmentprocesses into four categories: abrasion,erosion, sloughing and predator grazing.Abrasion and erosion both refer to theremoval of small groups of cells from thesurface of a biofilm. Sloughing, in contrast,refers to the detachment of large portionsof the biofilm, which may reach or be evenwider than the thickness of the biofilm itself(Morgenroth and Wilderer, 2000). It is alsolikely that detached, large biofilm clusterswill retain their protective propertiesagainst antimicrobials while in the liquidphase (Stoodley et al 2001).

1.1.3 Cultivation of biofilms

The potential of bacteria to form biofilmscan be measured in the laboratory usingmicrotiter plates (Danese et al 2000, Flet-cher 1990, Kolari et al 2001). The growthmedium and the bacterial inoculum aredispensed in the wells of the plate, andincubated at a chosen shaking rate andtemperature for a specific period of time.The wells are then emptied, washed andthe biofilm that has accumulated on thewalls of the wells is stained (e.g. crystalviolet). The colour intensity of the attachedcells can be measured by a plate reader.This method is simple, and allows a largenumber of analyses to be carried outsimultaneously. However, there are alsodrawbacks to this technique. Commerciallyavailable substrata (microtiter plates) arelimited to a number of different types ofpolystyrene. Furthermore, insufficientshaking during incubation or subsequentwashings may lead to sedimentation of thebacterial cells and false positive results.

When the development of biofilms needsto be followed on-line, flow cells (flowslides) connected to a CLSM can be used(Kuehn et al 1998, Wolfaardt et al 1994).In this technique, the bacteria form a

16 Introduction

biofilm under continuous flow conditionson the interior surfaces of a thin flowchamber, placed directly under themicroscope. Formation of the biofilm isvisualised by the attenuation of transmittedlight or fluorescence when bacteriacarrying the green fluorescent protein(GFP) are used (Martínez et al 1999, Kuehnet al 1998), or by staining the bacteria withnon-destructive, non-toxic fluorescentstains.Robbin’s devices are flow cells designedfor use in industrial applications (Ladd andCosterton 1990, Blanco et al 1996, Elverset al 1998). In this technique, process wateris allowed to flow through the device and abiofilm is formed on exchangeablesurfaces, so-called test coupons. The flowcells have to be sufficiently robust tooperate in industrial systems. Surfaces ofdifferent composition or quality and a rangeof flow rates can be used with Robbin’sdevices. The test coupons can be removedfrom the device and the biofilms formedon the coupons can be chemically orbiologically analysed.

A range of other types of equipment havealso been developed for biofilm studies: e.g.a rotating disc reactor (CBE, Montanareferred by Murga et al 2001) and Calgarybiofilm device (Ceri et al 1999), which ismodified 96 well microtiter plate system.

1.1.4 Methods for biofilmcharacterization

The confocal laser scanningmicroscope

The brilliant idea of combining a powerfullaser beam as a point light source and theuse of pinholes to cut off the non-focal lighthas completely changed the fluorescencemicroscopy of biological specimens. In aconfocal laser scanning microscope (CLSM),confocal pinholes in the light source (la-ser) and in the detector eliminate light fromabove and below the plane of focus. The

laser then scans in x-y directions throughthe specimen at different planes (z-direction) of focus. A 3-D image is builtfrom the information gathered by thedetector (Fig 5). As a result, each single x-y image is an optodigital thin section. Thethickness of each section depends on thelenses used and on the size of the confocalpinhole, and approaches the theoreticalresolution of the light microscope (≈ 0.2µm) (Pawley 1995b, Caldwell et al 1992).For example, when lenses with an NA of≤ 0.2 are used, the thickness of the opticalthin section is approximately 10 µm.However, this value rapidly drops to lessthan 1 µm when lenses with an NA of >0.6 are used. The use of laser beams alsomeans that the sample can be scanned inthe sagittal (zx) plane down to the depth ofthe laser light penetration, which isdependent on the opacity and opticalhomogeneity of the specimen (Centonzeand Pawley, 1995). Lawrence et al (1997)reported effective sectioning through 1 mmbiofilm material with 40 X waterimmersible lenses. If a cover glass is used,the thickness of the sectioned biofilm issmaller because the thickness of the coverglass is included in the sectioning distance.CLSM analysis requires no reagents otherthan aqueous solutions of fluorescent dyes,and can be performed on fully hydrated,living biomass. This is the basis forconstructing 3-D images of living biofilms.

The use of digital image collection allowsthe combination of fluorescence imagesfrom separate scans made at differentwavelengths of laser light. For instance,fluorescent phylogenetic probes can beused as tools for documenting the spatialdistribution of chosen species or genera ofbacteria, archaea or eukaryotic micro-organisms in a microbial community (Ste-wart et al 1997, Tolker-Nielsen and Molin,2000a). It also allows recognition of theEPS formed by the biofilm cells, as well asvisualization of the co-operation betweenseveral bacterial species within the biofilm

17Introduction

7. pinhole

laser

6. detectorphotomultiplier tube

3. specimen

1. tube lens, 2. objective,4. focal plane, 5. beam splitter

➞

Figure 5. Functioning of the confocal laser scanning microscope.A) A laser light beam is focused to a single point in the specimen. The pinhole (7) for excited lightdetermines the thickness of the focused layer (4); the smaller the pinhole, the thinner the focuslayer. During the scanning of the layer in focus the emission information is gathered by thedetector (6, photomultiplier tube) and transformed to an image on the computer screen. Aftercollection of information from several layers of focus, the data can be combined and 3 dimensionalpictures of specimen can be generated (modified from Zeiss Information with Jena Review, No. 5/95).B) CLSM image of the surface of paper stained with acridine orange. The panel on right side is acombination of 40 optical sections (0.5 µm). Bacteria are the small green dots on surface of woodfibre in the paper. The left panel is a 3-D stereo reconstruction of the same pile of optical sections.The differences in depth are more clearly visible. Measure bar is 10 µm.

A

B➞

18 Introduction

Table 1 Applications of confocal laser scanning microscopy for biofilm analysis

consortium (Manz et al 1999). The greenfluorescent protein (GFP) has been used forlocalizing chimeric proteins (Phillips2001). Rice et al (2000) studied “primary”biofilm formation and the progeny of thefirst generation of sessile cells, i.e. the“secondary” biofilm cells, by Pseudomonasaeruginosa PAO1 producing greenfluorescent protein. CLSM can also be usedto detect heavy metals. Wuertz et al (2000)used the complexing agent Newport Green,which fluoresces upon binding to Ni, Znor Co, to detect these metals in 3-D

bacterial biofilms. Table 1 compiles theapplications in which CLSM is frequentlyused. For more detailed information onnearly all aspects of confocal microscopy,the Handbook of Biological Confocal La-ser Microscopy (Pawley 1995a) is anexcellent source.

Study of biofilm microbialcommunities by in situ hybridization

Although in itself a powerful tool, thedevelopment of specific fluorescent nucleic

Target Tool Method Reference

Fluorescence in situ hybridization FISH

Detect bacteria belonging to a specific phylogenetic group

Fluorescently labelled oligo nucleotides

Amann et al 1992, Amann et al 1995, Zarda et al 1997, Hristova et al 2000, Amann and Schleifer 2001

Immuno-fluorescence

Detecting specific antigens on microbial cell surface

Fluorescently labelled mono- or polyconal antibodies

Bohlool and Schmidt 1980

CTC-DAPI, Live-Dead

Presence of exopoly-saccharide (EPS)

Analysis of EPS components

Fluorescently tagged lectins

This thesis

Biofilm architecture

Detection of pores and channels in biofilm

Fluorescent latex beads, Dextrans

Kolari et al 1998, Stoodley et al 1994, Lawrence et al 1994

Functional staining

Detection of vital functions of bacteria

Bredholt et al 1999, Maukonen et al 2000

Properties of biofilm

Presence and quality of bacteria

Nucleic acid staining

Distinguish bacteria from other attached substances

Syto 16, EtBr Kolari et al 1998

19Introduction

acid probes has added a high degree ofspecificity to CLSM. Such probes allow thedetection and identification of individualorganisms at levels from phylum rightdown to strain (Wimpenny 2000). The onlydisadvantage is that the fully hydratedbiofilms cannot be hybridized. The cellmembranes of the bacteria, forming thebiofilms, have to be broken in order to allowentry of the probes. The term in situhybridization implies that whole cells arehybridized and that the cells are viewed intheir natural microhabitat (Amann 1995).Different approaches using nucleic acidprobes are presented in Fig 6a. An earlierapproach based on immunofluorescenceproved that fluorescent signals allow high

resolution and fast detection byepifluorescence microscopy. Thefluorescently monolabeled, rRNA-targetedoligonucleotide probes were then shown toallow the detection of individual microbialcells (DeLong et al 1989). This madewhole-cell hybridization with rRNA probesa tool suitable for determinative,phylogenetic and environmental studies inmicrobiology (Amann et al 1990, Amannet al 1995, Amann and Schleifer 2001,Rabus et al 1996, Zarda et al 1997, Manzet al 1998, Kalmbach et al 2000, Hristovaet al 2000).

The method is based on the use of rRNA-targetted oligonucleotide probes. The

Figure 6. A) Flow chart showing the different options of using rRNA-targetted nucleic acid probesto detect the presence of different taxa in an environmental sample by in situ hybridizationtechniques (Amann et al. 1995).

20 Introduction

probes are usually 17 to 19 bases long. AsrRNA sequences are a patchwork ofevolutionary conserved regions, signaturesites can be identified, and probes designed,for any taxon between the level of thedomains Archaea, Bacteria, Eucarya andsingle species. Computer programs areavailable to generate probes based on thesequence data obtainable from publicdatabases (Amann and Schleifer 2001).Fuchs et al (1998) showed that the targetsite placement in the secondary structure

of rRNA strongly influences the intensityof the fluorescence emitted by the probe(Fig 6b). Several areas are so shaded thataccess of the hybridization probe to thetarget site during hybridization is hindered.

Scanning electron microscopy as atool to study biofilms

Scanning electron microscopy (SEM) hasfor long been used for determiningmicrometer-scale details from dehydrated

Figure 6 B) Distribution of relative fluorescence intensities of oligonucleotide probes, standardizedto that of the brightest probe, Eco1482, on a 16S rRNA secondary structure model. Differentcolours indicate different classes, decreasing in brightness from I (red) to VI (black) (Fuchs et al1998).

0-5% VI 6-20% V21-40% IV41-60% III61-80% II81-100% I

21Introduction

Table 2 Electron microscopic and atomic force microscopic methods for biofilm analysis

biological specimens. Chemical fixationand dehydration to allow imaging under ahigh vacuum, are needed to preserve thecellular structures. The final drying is inair or by critical point drying (Lounatmaaand Rantala 1991). As biological specimensconsist of light elements that are inefficientin scattering the electron beam, shadowingby vaporised platinum or gold is usuallycarried out in order to generate an electron-dense surface (Table 2). The steps performedprior to the microscopic inspection inevitably

modify the fine structure of the biologicalspecimens by causing shrinkage or byforming artefacts.

Energy dispersive X-ray analysis (EDS)for analysing elemental composition

Scanning electron microscopes arefrequently equipped with an energydispersive x-ray analyser. This equipmentpermits elemental analysis with a highhorizontal resolution of the inspected

tool method application referenceScanning electron microscopy SEM

Sample surface coated with electron dense material is scanned with an electron beam in vacuum

Distribution and morphology of bacteria

Väisänen et al 1998

Transmission electron microscopy TEM

Sample embedded in polymer resin is sliced to thin (0.1 to 2 mm) sections. Penetration of electron beam through the thin section is visualized

Distribution of bacteria and other substances in side of biofilm

Väisänen et al 1994, Kostyal 1998

Scanning probe microscopy SPM, Atomic force microscopy AFM

Mechanical sensor tip scans the surface topography of cells and measures the van der Waals forces towards or away from the tip

Nano resolution picture of living material

Dufrêne 2001, Telegdi et al. 1998, Razatos et al 1998, Kolari et al 2002

Field emission scanning electron microscopy FESEM

Field emission instead of normal wolfram emission allowing a narrower beam of electrons to be used.

High magnification (10000 x) imaging of bacteria distribution and morphology

Kolari et al 2001

tool method application referenceEnergy dispersive X-ray analysis EDS combined with SEM or TEM

Elemental characteristic energy changes are recorded as x-rays emitted by atoms in the specimen after electron bombing.

Elemental composition of specimen

Gharieb et al 1998, Sayer et al 1997, Otero et al 1997, Nurmiaho-Lassila et al 1990

Morphology of biofilms

Distribution of chemical elements in biofilms

22 Introduction

specimens. Changes in the electron energylevels (emitted x-rays) of the elements aretransformed to elemental distributioninformation (Lounatmaa and Rantala1991). The vertical depth of the analysis isdependent on the voltage used. Asbiological specimens are usually easilypermeable to electrons, signals from thesubstratum will also be detected. Lightelements like carbon, nitrogen, phosphorusand oxygen are important in biologicalstudies, and they can be detected with highquality analysers (Sivonen 1991).

2 Biofilm-driven processes

2.1 Biofouling

Biofouling is a general term referring to theundesirable accumulation of biotic depositson a surface. The deposit may containmicro- and macro-organisms (Characklis1991). Here we focus on microbial fouling,i.e. biofilms where microorganisms are themain biota in an organic film. Theorganisms may be embedded in a polymermatrix of their own making, or one froman external source. Biofilms may alsocontain non-microbial materials, such asprocess “dirt”. All material adhering to asteel surface hampers cleaning and mayserve as a substrate for microorganisms.Protozoa grazing on microorganisms mayalso occur in biofilms (Pedersen 1982).

How could biofouling be prevented inindustrial systems? Mechanical cleaning isthe oldest way to keep systems working.Usually the process needs to be shut downfor mechanical cleaning and the down timeis expensive. More modern cleaningmethods are CIP (cleaning in place) andthe use of biocides (Paulus 1993, Ross-moore 1995) to kill microbes present inprocess water (Wirtanen et al 1995) whilethe system is operating. However, theproblem with this approach is that biocides

are usually very effective against planktonicbacteria but less effective against biofilmbacteria (Brown and Gilbert 1993). Biocideefficiency can be improved by usingdispersants and other additives targetedagainst biofilm bacteria or their polymers.Physical means for preventing attachmentare high flow rates (> 1.5 m s-1, Geesey1993) and the use of “fouling releasesurfaces”, which rely on a low surfaceenergy to inhibit strong attachment (Ista etal 1999). Temperatures above > 60°C(Korkhaus et al 1996) have also been usedto attenuate microbial growth andbiofouling.

Measurement of the biomass attachedto surfaces

The degree of biofouling can be expressedas the quantity of biomass covering a givenarea of surface (Table 3). The metabolicactivity of the adhered microorganisms canalso be used to measure biofouling. Themost straightforward technique is todetermine the ATP content of the biomass(Fletcher 1990). ATP production is subjectto homeostatic regulation and an expensiveprocess for the cells. Therefore the contentof ATP in living microbial cells is relativelyconstant (ca. 0.02-0.1 % of dry weight,Salkinoja-Salonen 2002). ATP can beextracted from the cells and quantifiedusing the luciferase-luciferin enzymesystem of the firefly. The bioluminescenceis then measured by a luminometer. Theluminometer recording is proportional tothe concentration of ATP in the extract, andATP can be determined down to quantitiesof 0.01 to 0.1 fmol (Gregg 1991). As thebiofilm extract is likely to include inhibitorsof luminescence, an internal standard mustbe used. ATP measurement has been usedto replace viable cell counts and for theassessment of surface hygiene (Bretholt etal 1999, Maukonen et al 2000, Stone et al1999).

23Introduction

Table 3 Methods for estimating biomass quantities on non-living surfaces

direct Extraction of ATP by boiling in Tris EDTA and measuring the contents of ATP by light production using luciferase enzyme

This thesis

by swapping Swapping the surface with alginate swap and dissolving the ATP from the swap to be measured as light produced by luciferase enzyme

Davidson et al 1999

Gravimetric analysis

wet or dry weight

Substratum with biofilm is weighed

Plate count swapping and plating

The surface is swapped with alginate swap and bacteria are diluted to maximal recovery liquid and plated on agar. Colony forming units are counted

Carpén et al 1999

microscopy Bacteria are stained on surface and counted as number of cells or coverage of biofilm. Acridine orange (total biomass), CTC+DAPI (activity measurement)

Schaule and Flemming 1996, Fletcher 1990 Stewart et al 1994, Yu et al 1994

fluorometric measurement

Bacteria are stained on surface and counted by analyzer in units of emitted fluorescence

This thesis

Fluorescence

Assay method reference

Measure of biomass by content of ATP

24 Introduction

2

A

BH2S + 2e- H2 + S2- Fe2 + S2- FeS

H S + e- H + HS- Fe2+ + HS- FeS + H+

2.2 Biocorrosion

Microorganisms growing on a surfaceperform a variety of metabolic reactions,the products of which may promotedeterioration of the underlying substratum(Geesey 1991, Dowling and Guezennec1997) (Table 4). The main focus of thepresent thesis is laid on reactions where thesubstratum is of metal or of metal alloy.How does biocorrosion differ from “merecorrosion”? Flemming (1996) suggestedthat microorganisms do not introduce a“new” corrosion mechanism, but they canspeed up the chemical and electrochemicalcorrosion kinetics.Besides the fact that microorganisms andtheir metabolic products directly trigger

and accelerate corrosion processes, theirpresence in technical systems causes theformation of layers of variable thickness,generating cells with different oxygencontents (Fig. 7a). Such cells drive thedissolution of metals. It does not matterwhether the differential aeration cells weregenerated by microorganisms or by non-living deposits (Weber and Knopf 1996).In addition, bacterial metabolism mayproduce aggressive substances, oxidants,reductants, acids and complexing agents(Fig. 7b). This may lead to local patches oflow pH and/or redox potential underneaththe biofilm cluster and promote corrosionby solubilizing the metals by complexformation (Dowling and Guezennec 1997).

Figure 7. A) Localization of corrosive attack by concentration cells. B) Corrosion process in thepresence of sulfides. The arrows up and down show the anodic and cathodic partial reactions(modified from Heitz 1996).

25Introduction

Table 4 Bacterial taxa connected with corrosion

Methods for the detection of(bio)corrosion

The simplest method for determiningwhether surfaces in a process are corrodedor not is to take a sample, clean its surfaceand inspect it with a light and /or scanningelectron microscope. Removal of oxidesand other deposits before inspection isnecessary because the pits are frequentlysituated below the deposits (Tatnall 1991).The pits are commonly in the shape of apocket or crevice. The entrance on a steelsurface may be small, but the underlyingcavity may become large. Exchangeablecorrosion probes can also be used inprocess water systems to monitor forcorrosion (Tatnall 1991). When stainlesssteel is immersed in natural water the open

Genus or species reaction reference

Desulfovibrio desulfuricans 4H2+SO42-+2H+ -> H2S+4H2O Kuenen 1999

Desulfovibrio Corrosion of stainless steel pipes Otero et al 1997

Thiobacillus thiooxidans 2S0+3O2+2H2O -> 2SO 42-+4H+ Kuenen 1999

Thiocapsa Corrosion of stainless steel pipes Otero et al 1997

Thiobacillus ferrooxidans 4Fe+O 2+4H+ -> 4Fe 3++2H2O Kuenen 1999

Thiobacillus intermedius Corrosion of iron surface Telegdi et al 1998

Sphaerotilus sp. Corrosion of carbon steel Starosvetsky et al 2001

Gallionella Corrosion of stainless steel pipes Tatnall 1991

Siderocapsa Corrosion of stainless steel pipes Geesey 1991

Leptothrix sp. 2Mn2++O2+2H2O -> MnO 2+4H+ Kuenen 1999

Clostridium Hydrogen embrittlement Geesey 1991

Hydrogen forming bacteria

Sulphate and thiosulphate reducing bacteria

Sulphide and sulphur oxidizing bacteria

Iron oxidizing or reducing bacteria

Manganese oxidizing bacteria

circuit potential of the steel may increaseduring biofilm formation. This potentialshift towards the noble direction is calledennoblement, and indicates that corrosionmay subsequently follow (Anonymous1995, Scotto and Lai 1998, Carpén et al1995, 1997a, 1997b, Korkhaus et al 1996,Hakkarainen et al 1996a, 1996b).

26 Introduction

27

2 Aims

The specific aims of this study were to:

Determine the sequence of biological events that occur when a biofilmis formed on stainless steel during its ennoblement in Baltic Sea water.

Identify environmental factors that influence the biology-drivenennoblement of stainless steel in Baltic Sea water.

Develop a high fidelity laboratory simulator for the development ofmicrobial communities in the splash area of a paper machine. This wasdone in order to gather information on the salient features of the biologicalevents that are involved in deposit formation related to the corrosion ofprinting paper machine steels.

Analyse biofilms and their formation in the paper machine in situ, usingthe wet end area as an example.

Aims

I

II

III

IV

28

3 Materials and methods

The methods described in detail elsewhereare compiled in Table 5. The fluorescent stainsused in this study are listed in Table 6.

3.1 Stainless steel types

Stainless steel was used as the substratumfor biofilm growth in natural (Baltic Sea)or industrial (paper machine) waters. Thesteel types were those currently used forthe construction of machines andinstallations involving large-scale use ofwater, such as the cooling systems of powerplants, wet ends in the paper industry, andthe food industry (Table 7). Steel finishingsfrom delivery grade to 500 grit polishedwere used in the process equipment in these

industries. Biofouling and microbiallyinfluenced corrosion (MIC) wereinvestigated on steel types UNS S31600,S30400, S31254 and N08904 (Table 7).

3.2 Experimental set-ups used in thisbiofilm study

Biofouling and biocorrosion are majorcauses of downtime and maintenance costsin the paper industry and in power plants.Understanding the sequence of eventsleading to fouling and corrosion will helpto design protocols for minimizing theseproblems.

Method Described in:

ATP content of bacteria Paper II

In situ hybridization Kolari et al 1998, This thesis

SEM analysis Paper II, Väisänen et al 1998

TEM analysis Paper I, Nurmiaho-Lassila et al 1990

EDS elemental analysis Paper IV

CLSM analysis Paper II and Paper III

Open circuit potential measurement Paper I, Carpen et al 1995

Laboratory mesocosm Paper I

Splash zone simulator Paper IV, Carpen et al 1999

Flow cell for industrial applications Paper III

Staining with fluorescent beads Kolari et al 1998, Paper III

Staining with EtBr, syto 16, acridine orange, Live-Dead stains, concanavalin A

Kolari et al 1998

Scanning fluorometry Paper III

Table 5 Cataloque of the methods used in this study

Materials and methods

29

Table 6 Properties of fluorescent stains used in the CLSM and epifluorescence studies of thestainless steel grown biofilms.

Steel type

UNS C Si Mn S Cr Ni Mo Cu

S31600 0.03 0.56 1.51 0.04 16.9 10.7 2.6 -

S31254 0.01 0.42 0.40 0.002 20.1 17.2 6.1 0.67

S30400 0.04 0.46 1.48 0.005 18.4 8.6 0.1 0.2

N08904 0.02 0.53 1.48 0.002 19.5 25.0 4.5 1.4

Elements (%, w/w, elements other that Fe)

Table 7 Composition of the stainless steel types studies in this thesis.


Diameter or molecular weight

Fluorescence emission maxima

Concentration Surface properties

Nucleic acid stainsEthidium bromide 394 g mol -1 605 nm 100 µg ml-1 log Kow –0.38Acridine orange (DNA)

302 g mol -1 526 nm 100 µg ml-1 log Kow 1.24

SYTO™ 16 ≈ 450 g mol -1 518 nm 20 µM log Kow 1.48SYTO™ 9 (LiveDead®)

≈ 450 g mol -1 530 nm 3.34 µM

Propidium iodide (LiveDead®)

668 g mol -1 617 nm 20 µM

Carboxylate-modified

0.02 µm±15.8% 605 nm 9 x 10 12 beads ml -1 hydrophilic

Aldehyde-sulfate-modified

0.029 µm±20.1% 515 nm 3 x 10 12 beads ml -1 hydrophobic

Concanavalin A, tetramethylrhodamine conjugated

104 000 g mol -1 572 nm 200 µg ml-1 hydrophilic

Fluorescein 518 nmCy 3 563 nmOregon green 524 nm

Fluorescent beads for detecting porosity or surface properties

Miscellaneous stain for detecting glucose and mannose containing residues

Fluorescent labels used in oligonucleotide probes

Behavior of light in fluorescence techniques:Fluorescence is the result of a process thatoccurs in fluorophores or fluorescent dyes. Aphoton of energy (light) is supplied by anexternal source such as laser and absorbed byfluorophore. Due to this excitation the energylevel of fluorophore is changed. When returningto original energy level the fluorophore emits aphoton of energy (light). Due to energydissipation during the excited-state lifetime, theenergy of this emitted photon is lower i.e. longerwavelenght. The change in wavelenghts iscalled Stokes shift and is fundamental tofluorescent techniques (Haugland 1996).

When two different wavelengt of light areemitted at the same spot the visible light iscombination of the both. If same bacteria ishybridized both with fluorescein (green) andCy3 (red) labelled probes the join signal isyellow. The same happends with acridin orange,joint signal of DNA (green)and RNA (red). If red,green and blue are jointwe will see “normal”white light.

30

Baltic Sea water is used for coolingpurposes in power plants located along thecoast. For this reason, its capacity to formbiofilms or to induce biocorrosion ofstainless steels is not only of scientific, butalso of economic importance. We set up afield study site in the vicinity of Vallisaari,near the Helsinki coastline. The field studyequipment (details in Carpén et al 1995)was placed at the depth of 15 m, where thepenetration of daylight is < 1 % (estimatedbased on the method described by Vähä-talo 2000). This was done in order to reducealgal and other phototrophic growth on thesteel coupons.

For measurements that could not beperformed under field conditions, we builta large laboratory simulator of the BalticSea ecosystem, here called the mesocosm.It consisted of an illuminated chamber fedwith Baltic Sea water from a 800 lcontainer, and of a dark chamber fromwhich the water was pumped into Robbin’sdevices (Fig 1 in Paper I). Hydraulicretention times in the light (1850 lux, 12 hday-1, 100 l) and the dark (25 l) chambersof the mesocosm were adjusted to maintaina balance between the mineralization oforganic carbon (dark reaction) and theprimary production (light reaction). In thisway the concentration of dissolved organiccarbon was maintained at the natural levelof Baltic Sea water, 4-7 mg of non-particulated (< 0.02 mm) organic C perlitre. Stainless steel coupons were placedin modified Robbin’s devices submergedin the water recycling pipe connecting thedark and the light compartments of themesocosm. The mesocosm was fed withBaltic Sea water collected from the fieldstudy area. The composition of Baltic Seawater is given in Table 8. The Baltic Seabiofilms were grown for 4 to 8 weeks inthe field and in the laboratory during allseasons of the year. The conditions in thelaboratory mesocosm were similar to thenatural microbial ecosystem of Baltic Seawater, except for the temperature (23ºC).

The paper making industry continues to bea heavy user of fresh water in Finland (275million m3 per year in 2001, personalcommunication Kari Luukko, Metsäteolli-suus ry), in spite of the development inwater-saving technologies such as theclosure of mill water circulating systems.We investigated the biofouling of stainlesssteel surfaces inside the paper mill, as wellas in a laboratory model of the machinesplash area.

Biofilms were grown on coupons ofstainless steel (diameter 25 mm) in flowcells (Fig 1 in Paper III) connected to theside flow of the spray water circuit of apaper machine. The main properties of thespray water are shown in Table 8. In thepaper mill, spray water is used for cleaningthe returning wire(s) of the paper machine.The spray water is filtered and containsalmost no fibres.

A simulator was developed to studycorrosion related to biological depositformation in the splash areas of papermachines. It was modified from a testarrangement originally designed by Hak-karainen (1999) for studying the use ofchlorine dioxide for the bleaching of pulp.The simulator contained stainless steelcoupons, thermostated to 45ºC, and coveredby sheets of pulp under a glass dome. Thepulp sheets covering the coupon werecontinuously sprayed with simulated whitewater (Table 8) in order to mimick thesituation in the paper machine. However,the NaCl concentration was much higher(0.5 to 1 g l-1) than that in the “normal”white water of a paper machine (< 50 mg l-1).The set up of the simulator is shown inFigure 1 of Paper IV. The simulator enableddetermination of the mass balance betweenthe feed, the deposits formed inside thesimulator, and the effluent discharged fromthe simulator.

The formation of indigenous deposits onthe steels was studied using coupons of


31

Table 8 Characteristics of the waters used as substrate for biofilm growth in this thesis.

Baltic Sea water

Paper machine spray water

Simulated paper machine white water

pH 7.3-8.0 4.9-5.5 5.3

Temperature range C

0-32 41-48 45

SO4-2 mg l -1 340-540 1100 950

Cl- g l-1 2.6-3 0.022 0.5-1.5**

TOC mg l -1 4-7 270 400* (DOC)

Tot N µg l-1 300-800 260*

Tot P µg l-1 20-40 1600*

o

* Concentrations calculated from added amounts. Full recipe of simulatedpaper machine water is in M&M section of Paper IV.** The high amount of chlorine was added to promote corrosion events.

stainless steel (150 mm x150 mm) placedin the splash zone (≈40ºC) of the wiresection in a paper machine. The exposedcoupons were inspected by microscopyupon sampling.

3.3 In situ hybridization

Coupons of stainless steel with biofilmwere rinsed with sterile water, air–dried andstored desiccated at room temperature.Table 9 describes the oligonucleotideprobes used. The dry coupons werepretreated with ethanol: 30 % aqueousformaldehyde (90:10 v/v) for 5 min, rinsedwith sterile water, and the excess waterdrained off (protocol modified from Braun-Howland et al 1992). The ethanol-formaldehyde treatment was used to reduce

non-specific fluorescence (Amann et al1992). The hybridization mixture used forprobes EUB338, ALF1b and GAM42acontained 20-35 % formamide, 62.5 ng ofeach probe in 20 µl, 0.9 M NaCl, 20 mMTris-HCl pH 7.2, and 0.01 % SDS (Manzet al 1992). Twenty microliters of thismixture were applied on the biofilm on asteel coupon and incubated for 2 h at 46ºCin a humidity chamber. Unbound probe wasremoved with 2 ml of washing solution: 20mM Tris, 0.01% SDS, 5 mM EDTA and40-180 mM NaCl. The coupons were thenimmersed in 50 ml of the washing solutionat 48ºC for 20 min and rinsed briefly withsterile water. The hybridizations withSRB358 and DTM229 were conductedusing SRB specific conditions as describedby Zarda et al 1997.


32

Tabl

e 9

Olig

onuc

leot

ide

prob

es u

sed

for

in s

itu h

ybri

diza

tion

of b

iofil

ms.

Sequ

ence

NaC

l

5’ –

3’

(mM

) (w

ashi

ng)

16S

(338

-355

)

16S

alph

a(1

9-35

)P

rote

obac

teri

a

23S

gam

ma

(102

7-10

43)

Pro

teob

acte

ria

16S

delta

(385

-402

)P

rote

obac

teri

a

16S

(229

-247

)

* X

indi

cate

s th

at o

ne o

f ba

ses

A, C

, G o

r T

is u

sed.

308

Am

ann

et a

l 19

92, Z

arda

et

al 1

997

DT

M22

9A

AT

GG

GA

CG

CG

GA

X*C

CA

TD

esul

foto

mac

ulum

ge

nus

spec

ific

Ore

gon

gree

n15

-H

rist

ova

et a

l 20

00

SRB

385

CG

GC

GT

CG

CT

GC

GT

CA

GG

Cy3

20

180

Man

z et

al

1992

GA

M42

aG

CC

TT

CC

CA

CA

TC

GT

TT

Cy3

3540

Man

z et

al

1992

AL

F1b

CG

TT

CG

(C/T

)TC

TG

AG

CC

AG

Cy3

20%

form

amid

e (h

ybri

diza

-tio

n)

Ref

eren

ce

EU

B33

8G

CT

GC

CT

CC

CG

TA

GG

AG

Tdo

mai

n B

acte

ria

Fluo

res-

cein

2018

0M

anz

et a

l 19

92

Prob

eT

arge

t gen

e an

d E

.col

i nu

mbe

ring

Spec

ific

for

Lab

el


33

4 Results and discussion

4.1 Properties of biofilms as displayedby scanning electron microscopy, SEM

4.1.1 Visualising the formation ofbiofilms on stainless steel in the BalticSea water and in the paper machinewet end

Biofilms formed in the Baltic Sea watermesocosm (Fig 3 in Paper I) consisted ofclusters of different kinds of bacteria on theennobled stainless steel surface: bacteriawith an appearance similar to Seliberia(spiral), Caulobacter (stalked) orHyphomicrobium (cocci with tail), as wellas small rod-shaped bacteria (≤ 1 µm inlength, Fig 8a) (Boone et al 2001). Weregularly observed diatoms on the steelsurfaces in the immediate vicinity ofclusters of biofilm bacteria, and sometimesinside the cluster (Fig 11, Fig 6 in Paper I).

A conditioning layer is defined in theliterature as the film formed on solidsubstratum immediately after it isimmersed, before any bacteria becomeattached (Costerton and Lappin-Scott1995). Figure 8b shows an example of aBaltic Sea water biofilm that had grown for22 days in the mesocosm (≈23ºC). A thin(< 0.5 µm) film covering the attachedbacteria is visible. This film also covers thesteel surface in areas where there are noattached microbes (bottom left corner ofthe figure). The film may or may not bewhat has been described in the literature asa “conditioning layer”. Figure 8b showsseveral bacterial cells overlaid by a thinfilm, indicating that the film must have beengenerated after, and not before, theattachment of the primary colonizingbacteria. The film may originate from theexcreta of surface-attached bacteria or fromthe deposition of dissolved or colloidalmatter in the surrounding water (Kepkay

et al 1993, Schuster and Herndl 1995,Middelboe et al 1995), part of which maybe produced by the activity of the adheredbacteria (Wingender et al 1999).

Bacteria are known to excrete polymericorganic matter. Such polymers may play arole in the generation of biofilm clusters,or help individual bacteria to adhere. Theproduction of exocellular polymericsubstances is limited by the availableenergy. There should be a good reason forexcreting exopolymers, e.g. for protectionagainst grazing biota. Evidence obtained byFourier-transformed infrared spectroscopyof organic films in seawater indicates thatglycoproteins, proteins and possibly humicacids may be involved (Korber et al 1995).Humic substances are abundant in BalticSea water. This may in part explain the ori-gin of the organic films visible in Fig 8.Baltic Sea water is cold (0 to 17ºC,depending on the season) and oligotrophic.Seliberia, Caulobacter and Hyphomicrobiumare genera physiologically adapted to livein such environments (Corpe and Jensen1996, Stahl et al 1992, Lomans et al 1999).All three are oligotrophic chemo-organotrophs.

The warm (> 40ºC), mildly acid papermachine environment is very different fromthat of the Baltic Sea water, which is neutralin pH and colder. The total organic carboncontent of paper machine spray water isalso 50 times higher than that in Baltic Seawater (Table 8), and would therefore beexpected to favour bacteria different fromthose in the Baltic Sea. Biofilms grown for5 to 7 d in paper machine water flow in themill consisted of bacteria growing inclusters (Fig 9). No fungi or othereukaryotes were seen in the biofilm. Thebacteria were uniform in shape and sizewithin each biofilm cluster, but the shapesand/or sizes of the bacteria varied fromcluster to cluster (Fig 10). In the papermachine biofilms we observed only threemorphologically distinguishable types of

Results and discussion


Figure 8. Scanning electron micrographs of biofilm formed on stainless steel (UNSN08904) exposed to Baltic Sea water for 22 d in the laboratory mesocosm (≈23º C). A)Bacteria morphologically recognisable as; Seliberia (s), Caulobacter (c) andHyphomicrobium (h). Horizontal grooves originating from steel polishing are visible.B) A thin “conditioning-like” film covering large areas of the steel surface.

B

A

35

cell: uniformly coccoid (Fig 2 in Paper III),filaments with an identical appearance (≈50 µm in length, Fig 4 in Paper III), androds of varying length (1 µm to 6 µm, Fig5 in Paper III). Several long, rod-shapedbacteria have been earlier reported from thepaper machine wet end: the novel genusThermomonas (Busse et al 2002), speciesof Burkholderia (B. cepacia) Bacillus,Paenibacillus, Enterobacter and Meiothermus(Väisänen et al 1989a, 1989b, 1994 and1998, Kolari et al 2001, Suominen et al2002). Short rods detected in the papermachine wet end area have been identifiedas species of Sphingomonas, Acinetobacter,Klebsiella and Brevibacterium (Väisänen et al1994 and 1998). The cocci frequently foundin the paper machine wet end include thegenera Deinococcus, Cellulomonas andMicrococcus (Väisänen et al 1998, Kolariet al 2001).

4.1.2 Biofouling of stainless steelsurfaces exposed in the Baltic Sea and in apaper machine

The accumulation of material on non-livingsurfaces, whether formed by bacteria or bychemical and/or physical processes, iscalled fouling. Fouling decreases the flowrates in pipelines, decreases the efficiencyof heat transfer in heat exchangers(Flemming 1996), causes clogging, anddeteriorates the hygienic quality of thepaper product (Väisänen 1989a and 1989b,Pirttijärvi 2000). Biofouling thus results ina need for cleaning, thereby increasingcosts.

The Baltic Sea biofilms frequentlycontained cyanobacteria (Fig 11). Diatomswere more frequent on the steel surfacesexposed in the illuminated chamber of thelaboratory mesocosm than on the couponsplaced in the Robbin’s devices in the darkcompartment (Fig 17, Figs 6-8 in Paper II).

In the paper machine biofilms, wood fibresand pulping fines were the major non-

microbial components (Fig 12 in Paper III).Figure 12 shows a biofilm (5 days old) formedin situ in the spray water circuit of a papermachine. Spray water represents one of thecleanest types of water in a paper machine,and is low in fibre and contains primarilycolloidal and finely dispersed solids.Nevertheless, the figure shows a steelsurface covered mostly by substances otherthan bacteria. The bacteria were found tobe present in the first layer of the biofilm(Fig 7 in Paper III). The bacteria were nolonger visible by SEM during the laterstages of biofilm growth. The proportionof bacteria decreased as the age of thebiofilm increased (Fig 12 in Paper III). Theratio between bacteria and othercomponents in the biofilms did not remainconstant over the seasons. During springand autumn the nutrient content of naturalwaters is higher compared to that in otherseasons (Pesonen et al 1995). Bacteria usethe nutrients in the incoming water asgrowth substrates.

Conclusions on biofilm formation asobserved by SEM

The scanning electron microscopicevidence in our study shows that the biofilmbacteria in the two environments, i.e. theBaltic Sea and a paper machine, weremorphologically different. The differenceswere presumably related to the differentselective conditions in these environments.Other types of material were primarilyattached on top of the thin layer of primary-attaching bacteria in both the Baltic Seawater and paper mill biofilms.

4.1.3 Corrosion- related microbiologialevents on stainless steel in Baltic Seawater: factors influencing ennoblement

Our aim was to identify the environmentalor microbiological factors influencing thedevelopment of ennoblement, and therebythe risk of corrosion of the steel during thegrowth of biofilms.


36

400 x

100 x


Figure 9. Scanning electron micrographs (SEM) of biofilm formed on stainless steel(UNS S31600) during 5 days of exposure to paper machine spray water (≈45º C) flowin the mill. The magnifications are 100 x, 400 x, 800 x and 3000 x. The box shows thearea used for the zoom-up. The low magnification micrograph (100 x) shows the weblike structure of the biofilm.

37

AAAAA

BBBBB


The 400 x to 3000 x magnifications show differently sized rod shaped bacteria (2-7µm). SEM technique requires the samples to be fixed (glutaraldehyde) and fully driedbefore the image is taken. The desiccation may cause the microbes to shrink down to1/3 or 1/2 of their volume compared to living cells.

800 x

3000 x

38

A

B


Figure 10. Scanning electron micrographs of 2 different biofilm clusters formed on acoupon of stainless steel (UNS S31600) exposed to paper machine spray water flowfor 5 days of in the mill. Panel A shows biofilm aggregates containing bacteria withuniform cocci like appearance. Panel B shows rod shaped bacteria of 2 to 1 µm inlength and large cyanobacteria like cells (3 µm x 8 µm).

39

A

B


Figure 11. Scanning electron micrograph of biofilm formed on stainless steel (UNSS31254) exposed to Baltic Sea water for 22 d in the laboratory mesocosm (≈23ºC).Panel A shows a Vorticella like protozoan grazing in the biofilm. Panel B shows achain of cells resembling the cyanobacterium Anabaena and pieces of Crustaceans .

40

Ennoblement of the steel was inhibited bydoubling the concentrations of ammoniumand phosphate (increase of 778 mg N l-1

and of 33 mg P l-1) in the Baltic Sea waterin the laboratory mesocosm underotherwise favourable conditions. Noennoblement occurred in the steel couponsin the laboratory ecosystem when the feedwas Baltic Sea water enriched by anaccidental, month-long discharge of treatedsewage, which raised the total nitrogencontent in the seawater two fold over thenatural level of 300-800 mg N l-1, butcaused no changes in the phosphorus level.The biofilm that grew in this water differedfrom the biofilm formed in the moreoligotrophic water in that it was easilydetached from the steel surface and had nocompact microbial clusters (Fig 8 in PaperII).

The conditions that allowed formation ofennobling biofilms were a flow rate of 10-35 mm s-1 and low nutrient concentrations(Table 1 in Paper II). When the flow ratewas low, i.e. < 10 mm s-1, loosely attached,fluffy biofilms grew on the steel surfaces.These biofilms had a large surface coveragebut did not increase the open circuitpotential of the underlying steel (Tables 1,2 in Paper II).

In the highly structured biofilm, thebacterial clusters that co-emerged withennoblement of the steel were separatedfrom each other (clear-edged structure). Ahigh flow rate favoured the formation of abiofilm in which the microbial cluster wasstrongly attached to the substratum (the lastparagraph in the Results Section of PaperI, and the last paragraph in the Results


Figure 12. Scanning electron micrographs of a biofilm formed on stainless steel (UNSS31600) during exposure to paper machine spray water for 5 days in the mill. Herethe biofouled steel is covered by fines and fibers (F) originating from pulping. Bacteria(B) are minor constituents in this biofilm.

41

Section of Paper II). A compact biofilmcluster may prevent oxygen penetrating intothe steel surface below the biofilm. As aconsequence, the flow of electrons maybecome channelled towards iron (Fe3+–>Fe2+

[E0

´+770 mV], O2 –>H

2O [E

0´+820 mV],

Nealson and Myers 1992). These structuralproperties will contribute to the formationof differentiated aeration cells, leading tocorrosive events (Watkins Borenstein1994).

During the winter season the open circuitpotentials of the stainless steel couponsimmersed in the Baltic Sea increased from-200 mV

sce to 250 mV

sce in 50 days at 0°C.

Coupons of the same steel showed noennoblement at all within 38 days whenimmersed in Baltic Sea water thermostatedat 32 to 36 °C in the laboratory mesocosm.When the temperature of the water wherethese coupons were immersed was cooledto 26°C, ennoblement occurred within 20days. This shows the existence of an uppertemperature threshold of ≤ 32°C for theennoblement of stainless steel in Baltic Seawater (Fig 2 in Paper II). The temperaturedependence of ennoblement indicates aconnection to temperature-limited biology.Since the temperature in the Baltic Sea istypically between 0 to 15°C (Pesonen et al1995), an increase of 20°C may block thegrowth of the indigenous microorganismsinvolved in ennoblement.

Microscopic inspection showed that thesteel surface was only partially coveredwith biofilm at the time when ennoblementoccurred in the laboratory mesocosm withBaltic Sea water (Fig 4 in Paper I, Fig 6 inPaper II). Partial coverage indicates thatdissolved oxygen had free access to theuncovered parts of the metal surface duringennoblement. The ennoblement of stainlesssteel may thus require partial, but not fullcoverage, of the steel surface by a compactbiofilm that insulates patches of the steelsurface against the diffusion of oxygen.Ennoblement may be associated with the

co-existence of patches with differentaccess to oxygen. The patchy, electricallyinsulating biofilm may thus form the basisfor the onset of corrosion.

We measured the content of ATP in biofilmson the steel surface, as well as the total dryweight, in order to determine whether thereis a threshold for the density or the surfacecoverage of biofilm before ennoblementcan occur. If this was the case, such aparameter could be used for predicting theoccurrence of corrosive events.Our results showed that neither the amountof ATP nor that of dry matter in Baltic Seabiofilms predicted the onset ofennoblement and thus a potential risk forcorrosion of the steel (Table 2 in Paper II).The results were compatible with the viewthat corrosive events are related to athreshold in the surface coverage of aspecific type of biofilm rather than to theoverall biomass.

4.1.4 Biological events linked to thecorrosion of stainless steel in a papermachine

Only a few studies have been carried outon the microbial communities in the wetend and the splash area of paper machines(Väisänen et al 1989a, 1989b, 1994 and1998, Pirttijärvi et al 2000, Lindberg et al2001b, Kolari et al 2001, Busse et al 2002),or on microbially induced corrosion inthese sites (Vestola and Korhonen 1976,Soimajärvi et al 1978). The wet end areaof a paper machine frequently suffers fromcorrosion (Carpén et al 2001a, Thorpe1985). We found high concentrations ofsoluble oxalates plus oxalic acid (max 2300mg kg-1), and of thiosulphate (max 16000mg kg-1), in wet end deposits. Thiosulphateis known to be corrosive to stainless steel(Garner and Newman 1991, Laitinen 1999).The oxalic acid present under biofilms andother deposits on steel surfaces may lowerthe pH down to 2, thereby contributing tocorrosion (Anonymous 1994). The process


42

water contained practically none of thesecompounds (Table 2 in Paper IV): < 5 mgthiosulphate kg-1 wet wt. and < 2 mg oxalatekg-1.

A laboratory simulator was designed toinvestigate the origins of oxalate andthiosulphate in the wet end deposits ofpaper machines (Paper IV, Carpen et al1999, 2000, 2001abcd). In this setup, called“the splash area simulator”, stainless steelcan be exposed to process water underconditions simulating those in the splashareas of the paper machine wet end.

The splash area simulator was used toinvestigate the events leading to theformation of corrosive deposits in the papermachine. In the simulator, a coupon ofstainless steel was overlaid with sheets ofpulp preinoculated with strains of Bacillussubtilis, Desulfovibrio desulfuricans andDesulfotomaculum thermoacetoxidans, andwith an unidentified SRB culture enrichedfrom paper machine white water (Paper IV,M&M). The simulator was fed withsynthetic white water (Table 8), spiked withvarying amounts of chlorine, to speed upcorrosion. The chemical composition of thewater leaving the simulator was measured.Thiosulphate (80 mg l-1) and oxalate (70mg l-1) appeared in the effluent leaving thesimulator when the simulator had beenoperating for two weeks. As these ions werenot present (thiosulphate < 10 mg l-1,oxalate < 5 mg l-1) in the feed, thiosulphateand oxalate must have been formed de novoinside the simulator.

The open circuit potentials of the steelcoupons placed in the paper machinesimulator did not increase, indicating thatno ennoblement occurred. After 28 days ofoperation of the simulator, the stainlesssteel coupons were removed and inspectedby SEM for pitting. Stainless steel S30400had become pitted over the whole surfaceunderneath the pulp sheet during exposure.

Stainless steel S31600 showed pits only inthe peripheral area, i.e. at the junction ofthe pulp sheets and the glass dome coveringthe sheets. Microbial cells of 2-3 µm in sizewere found in the pits (Fig 13a) formed onsteel S30400 during exposure in thesimulator. These microbes may havecontributed to the corrosion of the steel, orthey may simply have sought shelter in thealready formed corrosion pits. The pitswere found in the preinoculated simulatorsonly. Sulphate reducing bacteria weredetected on steel surfaces in thepreinoculated simulators (3rd chapter ofResults in Paper IV).

Using SEM-EDS we found crystals on thesteel surfaces exposed in the simulatorwhere the pulp sheets had beenpreinoculated. The crystals had anelemental composition characteristic ofcalcium oxalate, calcium carbonate andcalcium sulphate (Fig 13bcd). Crystallizedcalcium oxalate was also found in thedeposits collected from the interior wallsof a pulp slurry storage tank of a papermachine. Calcium oxalate is insoluble inwater. If such crystals enter the wet end areaof a machine, the pick-up felt and the wiremay become clogged.

With SEM-EDS we detected deposits witha high silicon content all over the steelsurfaces that had been exposed in thepreinoculated laboratory simulator.Stoecker and Pope (1993) and Otero et al(1997) may have observed a similarphenomenon on steel exposed to a coolingwater system using seawater. Theysuggested that the accumulation of silicon-containing materials on the stainless steelwas linked to microbially inducedcorrosion. The same authors also found thatbacteria belonging to the sulphate-reducinggenus Desulfovibrio and the sulphide-oxidizing genus Thiocapsa were presentduring the microbially induced corrosionevents in the coastal sea water. Our splash


43

area simulators were inoculated with Des-ulfovibrio desulfuricans (DSM 642),Desulfotomaculum thermoacetoxidans(DSM 5813) and a mixture of unidentifiedsulphate-reducing bacteria from the papermachine. Metabolism similar to that notedby Stoecker and Pope (1993) and Otero etal (1997) may have also taken place in thesplash area simulator. According to Bert-helin (1983) large amounts of complexingagents, like oxalic acid, can promote thesolubilization of mineral elements (Si, Al,Fe, Mn, Ca, K) from granite rocks. In oursystem the silicon-containing deposits mayhave originated from the etching of silicon-containing steel (Table 7) in the aggressivelyacid conditions generated by the largeamounts of oxalic acid (Table 2 in Paper


IV). Microbiological transformations ofsilicon compounds were reviewed byKrumbein (1983).

Silicon compounds are excellent electricalinsulators. Therefore silicon-containinglayers may lead to the development ofdifferential charges on the stainless steel,strengthening the theory that partialcoverage by insulating material may beinvolved. Coverage of the steel surface bymicroorganisms or other material to theextent that it changes the access of oxygento the steel surface, leads to a differentialaeration cell. Formation of differentialaeration cells is considered to be one of themajor events in microbially influencedcorrosion (Heitz 1996).

44

A

B


Figure 13. Scanning electron micrographs of stainless steel coupons (UNS S31600)pitted after being exposed to simulated paper machine process water in a splash areasimulator operated at ≈45ºC for 28 d. A) Yeast-sized microbial cells are visible coveringa pit. B) The EDS analysis revealed that this cubic crystal consisted of calcium andcarbon, most likely of calcium carbonate or calcium oxalate.

45

C

D


Figure 13 C) An example of an area where crystals covered the steel surface and thecrevices formed in the steel. D) Crystals with x-form resembled calcium sulphate inelemental composition based on EDS analysis.

46

4.2 Properties of biofilms on stainlesssteels as displayed by confocal laserscanning microscopy (CLSM)

4.2.1 In situ hybridization in a studyof biofilm forming bacteria

Fluorescent in situ hybridization withlabelled oligonucleotide probes was usedto obtain information on the taxonsconstituting the microbial biofilmsgenerated in the spray water area of a papermachine and during the process ofennoblement of stainless steel in seawater.

We used eubacterial oligonucleotide probeEUB338 to detect the presence of bacteriaand to test the success of the appliedhybridization procedure. The hybridizationprotocols reported in the literature are forbiofilms on the surfaces of glass or plastic.Stainless steel grown biofilms have notbeen investigated by FISH prior to thisstudy. We found that EUB338 hybridizedmost of the bacteria (green fluorescence,fluorescein) in the Baltic Sea biofilms (Fig14b), proving that the chosen hybridizationprotocol was operative. The alphaProteobacteria-targeted probe ALF1b (red,Cy3) was used together with probeEUB338 (green) in Figures 14b and 15. Thetwo probes hybridized into separatebacterial clusters in the biofilm. This isnoteworthy, because the alphaProteobacteria group should form a partof the domain bacteria to which EUB338is targeted. The double-labelled bacteriawould have fluoresced yellow (joint greenand red). 11.8% of the bacterial 16S rRNAsequences in the databases matchedperfectly with the ALF1b probe, while theyexhibited 1 to 2 mismatches with theEUB338 probe (Kolari et al 1998). The redfluorescence is explained if the bacteria inthat cluster belong to the 11.8 %mismatching with EUB338. The gammagroup of Proteobacteria was searched forusing the GAM42a probe (red fluorescence,Cy3) together with the EUB338 probe

(green). The results showed that only aminority(< 15 %) of the biofilm-formingbacteria in the Baltic Sea biofilms weremembers of the gamma group ofProteobacteria matching with GAM42a(Fig 14a). This is less than the score forALF1b, which hybridized with 25-30 % ofthe biofilm bacteria, indicating they weremembers of the alfa group ofProteobacteria.

Based on the results, above we presentlyconclude that the majority of the ennobledstainless steel biofilm-forming bacteria inthe Baltic Sea were recognized by theEUB388 probe, but not by the alpha or thegamma proteobacterial probes. Thereforethe bulk of the biofilm bacteria mayrepresent eubacteria other than the alphaor gamma Proteobacteria. SEMmicrographs of the Baltic Sea biofilms (Fig8) showed the presence of Caulobacter,Hyphomicrobium and Seliberia-likebacteria, all of which belong to alphaProteobacteria. Based on the sequence dataalone, the Caulobacter should havehybridized with both probes (EUB338 andALF1b), whereas Hyphomicrobiummatched with EUB338 only and Seliberiadid not match with EUB338 or with ALF1b.

Paper machine splash area biofilms fromthe mill were hybridized to the EUB388probe (fluorescein, green) targeted to thedomain bacteria, and to the SRB385 probe(red, Cy3) targeted to the sulphate-reducingdelta subclass of Proteobacteria. Bacteriawith a positive response to EUB388 (green,fluorescence) were detected, but if the redfluorescence expected from SRB385 waspresent at all it was masked by the strongautofluorescence of the paper machinefines (Fig 16b).

Probes for the gram-positive genusDesulfotomaculum DTM229 (oregongreen) and for the sulphate-reducing deltaProteobacteria SRB385 (red, Cy3) wereused to determine the presence of bacterial


47

groups with a metabolic capacity to reducesulphate into sulphur-containing ions oflower oxidation states. Paper machinewater contains 0.5 to 1 g of SO

42- l-1 (Table

8), and sulphate is the substrate forsulphate-reducing bacteria. Their products,reduced sulphur-containing ions, maycause corrosion of stainless steel (Hamil-ton and Lee 1995). The splash areabiofilms, formed on stainless steel in thepaper machine, showed a faint binding byDTM229 to large sized bacteria (Fig 16a).No red fluorescing cells were detected,even though the hybridization conditionswere chosen to favour SRB385 (Table 9).

Oligonucleotide probes EUB338 andALF1b belong to fluorescence brightnessgroup III described by Fuchs et al (1998),based on the secondary structure of rRNA.This means that they fluoresce with anintensity of 44 to 58% of the intensityexhibited by the brightest probe, Eco148L.The fluorescence intensities of EUB338and ALF1a are therefore expectedly twotimes higher than that of the group IVprobes SRB385 (≈39%) and DTM229(≈26%). This may explain the observeddifferences in the signal detection ofDTM229 versus EUB338 (Fig 16).

The result indicates that deltaproteobacteria were not major biofilmconstituents in the splash area of the papermachine, and that the genusDesulfotomaculum was present but notabundant. The Baltic Sea water biofilmswere more readily analysable than the papermachine biofilms using hybridizationmethods, due to less interference byautofluorescencing substances. In caseswhere autofluorescence was observed inthe Baltic Sea water biofilm it was of adefined shape (diatoms) and thereforeeasily identified. The large amounts ofautofluorescencing papermaking fines inthe biofilms disturbed detailed,fluorescence-based detection systems likein situ hybridisation. The presence of

soluble and insoluble salts (Figs 13bcd,Table 2 in Paper IV) are also likely todisturb the hybridization procedure.

4.2.2 Use of autofuorescence andspecific fluorescent stains foranalysing the biofilms formed onstainless steel

The algae, especially silicon-containingdiatoms, present in Baltic Sea biofilms havestrong red autofluorescence. One or twodiatoms per microscopic field (200 µm x200 µm) were regularly observed on thesurface of the stainless steel in thesebiofilms. Diatoms were useful for theinterpretation of other CLSM resultsbecause they served as indicators of thelaser beam penetration through the biofilm(Fig 6 in Paper I, Fig 7 in Paper II).

We used tetramethylrhodamine conjugatedconcanavalin A (conA) as an indicator ofthe presence of glucose- or mannose-containing polysaccharides in the papermachine biofilms. It stained yeast-sizedcells, but not as successfully as the smaller(Ø 1 µm) biofilm-forming bacteria (Fig 3in Paper III). The conA conjugate stainedthe amorphous fines of paper machineslimes, most probably due to the presenceof mannose and glucose in cellulose andhemicellulose, respectively. Yeast-sizedcells were found (Fig 13a) on steel couponsleft for 28 days in the paper machine splasharea simulator. Their cell size and shapewas similar to the yeast-sized cells stainedwith con A (Fig 3 in Paper III).

4.2.3 Properties of the biofilmclusters grown on stainless steel

Knowledge of biofilm porosity is importantfor the planning of trategies for disinfectionand for the removal of biofilms. If thebiofilm cluster remains porous whilegrowing in height/thickness, cleaningsolutions may penetrate to the roots of the


48

AB


Fig

ure

14.

CLS

M i

ma g

es o

f a

biof

ilm o

n a

coup

on o

f st

ainl

ess

stee

l (U

NS

S316

00),

im

mer

sed

for

30 d

ays

in B

altic

Sea

wat

er i

n a

labo

r ato

rym

esoc

osm

(≈2

3º C

). T

he im

a ges

ar e

pile

ups

of 2

0 op

tical

sec

tions

, 0.5

µm e

ach.

A) T

he im

a ge

show

s a

biof

ilm h

ybri

dize

d w

ith th

e pr

obe

EU

B33

8(g

reen

, tar

gete

d fo

r th

e do

mai

n ba

cter

ia)

and

with

GA

M42

a (r

ed, t

arge

ted

for

the

gam

ma

grou

p of

Pro

teob

acte

ria )

. B) T

he im

a ge

show

s a

biof

ilmhy

brid

ised

with

the

prob

e E

UB

338

(gr e

en, t

arge

ted

for

the

dom

ain

bact

eria

) and

with

ALF

1b (r

ed, t

arge

ted

to a

lpha

gr o

up o

f Pro

teob

acte

ria )

. The

mea

sur e

bar

s ar

e in

µm

.

49

AB


Fig

ure

15. C

LSM

ima g

es o

f a b

iof il

m o

n a

coup

on o

f sta

inle

ss s

teel

(UN

S S3

1600

), g

r ow

n fo

r 30

day

s in

Bal

tic S

ea w

ater

in a

labo

r ato

ry m

esoc

osm

(flo

w s

peed

≈30

mm

s-1, ≈

23º C

). T

he im

a ges

ar e

pile

ups

of 2

0 op

tical

sec

tions

, of 0

.5 µ

m e

ach.

A) T

he b

iof il

m h

ybri

dize

d w

ith th

e pr

obe

EU

B33

8(g

reen

, tar

gete

d fo

r th

e do

mai

n ba

cter

ia) a

nd w

ith p

r obe

ALF

1b (r

ed, t

arge

ted

for

alph

a gr

oup

of P

rote

obac

teri

a ). B

) A z

oom

up

of th

e ce

ntr a

l ar e

am

ark e

d in

pan

el A

. Pan

els

A a

nd B

sho

w th

at b

acte

ria

with

gr e

en fl

uor e

scen

t lab

els

(EU

B33

8) w

ere

a m

ajor

ity b

ut b

acte

ria

with

red

(ALF

1b)

and

yello

w (

doub

le la

bele

d w

ith E

UB

338

and

ALF

1b)

fluor

esce

nce

wer

e al

so s

een.

The

mea

sur e

bar

s ar

e in

µm

.

50

AB


Fig

ure

16. C

LSM

ima g

es o

f a b

iof il

m o

n a

coup

on o

f sta

inle

ss s

teel

(UN

S S3

1600

) exp

osed

for

50 d

ays

in th

e sp

lash

ar e

a of

the

wir

e se

ctio

n of

apr

intin

g pa

per

mac

hine

. A)

A b

iof il

m h

ybri

dize

d w

ith th

e pr

obe

SRB

358

(red

, tar

gete

d to

the

delta

gr o

up o

f Pro

teob

acte

ria )

and

with

DTM

229

(gre

en, t

arge

ted

to th

e gr

am p

ositi

ve g

enus

Des

ulfo

tom

acul

um).

B) A

bio

f ilm

hyb

ridi

zed

with

the

SRB

358

(red

) and

the

EU

B33

8 (g

r een

, tar

gete

dto

the

dom

ain

bact

eria

) pr o

bes.

A g

r oup

of g

r een

fluo

r esc

ing

bact

eria

is v

isib

le a

mon

g th

e r e

d au

toflu

ores

cing

f ine

s an

d f ib

res.

Mea

sur e

bar

s ar

ein

µm

.

51

cell clusters and kill the cells forming theplatform of the whole cluster. We usedfluorescent stains of different molecularsizes and of different surface properties toassess the permeability properties of thebiofilm clusters. As a tool to assess biofilmporosity, we used fluorescently labeledlatex beads with a diameter of ≈0.02 µm,i.e. smaller than any of the bacteria in thebiofilms. The purpose was to detectwhether there were flow channels inside thebiofilm and to determine the generalporosity of the biofilm clusters.

The biofilm clusters grown during theennoblement of stainless steel in Baltic Seawater were poorly (≤10 cell layers, PaperII) penetrable to concanavalin A(hydrophilic stain, Table 6) and also tostains (Table 6) with different polarityproperties: acridine orange (log K

ow 1.24,

fig 17), ethidium bromide (log Kow

-0.38)and SYTO™ 16 (log K

ow 1.48) (Kolari et al

1998). All these substances were excludedfrom the cores of the biofilm clusters.Autofluorescing diatoms were visible (Fig17, Fig 6 in Paper I) underneath the biofilmclusters, indicating that the laser beam hadpenetrated down to the steel surface. Theresults support the view that biomass witha high diffusion barrier but goodpenetration of the laser beam (Fig 7 andFig 6f in Paper II) was present inside thebiofilm clusters on the ennobled steel. Suchbiomass may be involved in theennoblement of stainless steels.

We found that the fluorescent beads of 0.02µm diameter with hydrophilic orhydrophobic surfaces only weaklypenetrated the biofilm grown on stainlesssteel in situ under conditions of high flow(1.8 m s-1) in paper machine spray waterfor 7 days (Fig 2 in Paper III). The beadspenetrated through the uppermost 2 to 3 celllayers in the biofilms with a total thicknessof up to 30 µm. In contrast to the beads,the stains SYTO™16, SYTO™9 (Fig.18a), ethidium bromide (Fig 18b) and

propidium iodide (Fig 18a) penetrated intothe core space (through 25 to 30 cell layers)of the same biofilm clusters. Figure 19shows a single optical section at 3 to 9 µmdistance from the steel surface of a biofilmformed in the splash area of the papermachine. It shows that the papermakingfines are a major ingredient in the structureof this biofilm. The bacteria (red, stainedwith propidium iodide) formed cell clustersamidst the fines of pulping origin. As thebiofilm was dry at the time of sampling inthe paper machine, this may explain whyall the bacteria appear dead (redfluorescence).

Baltic Sea biofilms were grown in thelaboratory ecosystem under a flow of 30mm s-1 at 23 ºC. The paper machine biofilmswere grown under 60 times higher flowrates and at a temperature of 40 to 50ºC.The exposure periods in the Baltic Sea werelong, up to 76 days, but in the papermachine only 7 days. The dyes penetratedinto the biofilm core in the paper machinegrown biofilms. This may be due to thethinner biofilm. The Baltic Sea grownbiofilms were older and thicker (up to 120µm). The high cluster depth of the BalticSea biofilms may have limited thediffusion. Figure 17b shows a pile up ofoptical z sections of a 120 µm thick BalticSea biofilm where acridine orange hadpenetrated down to a depth of 10 to 25 µmfrom the surface of the biofilm. Thepenetration depth of the dyes (ethidiumbromide, acridine orange and SYTO™16)was thus similar in the paper machine andin the Baltic Sea water biofilms, indicatinga similar penetration resistance. The limitedthickness of the paper machine biofilmsallowed penetration through the totalthickness of the biofilm.

A high flow rate (> 1.5 m s-1) has beenconsidered to protect against fouling andagainst microbially induced corrosion(MIC) (Geesey 1993). Stoodley et al (1999)found that a flow velocity of 0.7 to 1.3 m s-1


52

limited the thickness of the biofilm formedon glass in minimal salts medium withglucose as the carbon source to ≈20 µm.Our results showed that a high flow rate(1.8 m s-1) limited the biofilm thickness(max 30 mm, exposure time up to 12 d).According to Vieira and Melo (1999), thethin biofilms formed under high flow ratesshould be easily penetrated by the biocidesused for limiting biofilm growth due to theshort diffusion distance.

Conclusions about biofilm formationobserved by CLSM

The exclusion of stains and latex beads of0.02 µm in diameter by the Baltic Seabiofilm indicates that a barrier exists insidethe cell clusters grown during theennoblement of steel. The stains with asmall molecule size penetrated down to ≈20cell layers, but did not reach the steelsurface. The high penetration resistance ofthe biofilm may also be valid for oxygen,leading to the formation of differentialaeration cells. Such a situation wouldpermit anaerobic metabolism in the deeperlayers of the cluster, closest to the stainlesssteel surface. The formation of differentialaeration cells paves the way for thegeneration of anaerobic metabolites, suchas reduced sulphur-containing anions.Differential aeration may increase the opencircuit potential of steels, leading toennoblement and, subsequently, tocorrosion of the steels.Fluorescent stains penetrated throughoutthe biofilms grown in a high flow in papermachine water. The mechanisms of steelcorrosion in paper machines may bedifferent from those in Baltic Sea water.

4.3 Methods developed in this study

We developed, through trial and error, anovel type of flow cell for use ininvestigating the biofouling of steels in an

industrial environment. The design isshown in Figure 1 of Paper III. The sidepanels of the flow cells were made ofpolyacryl, and the top and bottom panelsof stainless steel. Sample holders for testingwere fitted into the top and bottom panels.The transparent polyacrylic side panelsserved to allow visual inspection of the flowcell interior without having to remove thesamples.

The flow cell was directly connected to theside flow (average 1.8 m s-1) of the watercirculation system of a paper machine andoperated under machine pressure, with noneed for pumps. The flow cell operatedcontinuously for up to 3 weeks withoutremoval for cleaning. The structure of theflow cell allowed separation of the top andbottom panels from the sidewalls forintensive cleaning between experiments.The exchangeable coupons of stainlesssteel made it possible to compare thebiofouling tendency of different types ofsteel and of different grades of surfacefinishing. The flow contact surface of thesteel coupons had a diameter of 25 mm.Up to 16 coupons could be placed in eachflow cell. This made time-lapse studiespossible. The flow cells thus proved to beefficient tools for determining therelationship between steel surfacesmoothness and the tendency to biofoul ina paper machine. The results demonstratedthat 500 grit steel was less sensitive tobiofouling compared to 100 grit steel (Fig9 and 10 in Paper III).

Scanning fluorometers are available on themarket for monitoring microtiter plates. Wedevised a protocol for using this instrumentfor direct scans of the biofilms on stainlesssteel. To accomplish this, the biofilms onthe steel coupons were stained with afluorescent dye. The coupons were thenplaced in the wells of microtiterplates, onecoupon in each of the 6 wells. The scanningarea of the fluorometer was programmed


53Results and discussion

Fig

ure

17. C

onfo

cal l

aser

sca

nnin

g m

icr o

scop

y (C

LSM

) im

a ges

of a

n ac

ridi

ne o

r ang

e st

aine

d bi

ofilm

on

a st

ainl

ess

stee

l (U

NS

S316

00)

coup

ongr

own

for

76 d

ays

in B

altic

Sea

wat

er in

the

labo

r ato

ry m

esoc

osm

(ver

y lo

w fl

ow s

peed

, ≈23

ºC).

A) 3

D s

ter e

o im

a ge

of a

pile

up

of 2

50 in

divi

dual

optic

al s

ectio

ns (0

.5 µ

m th

ick

eac h

) of a

bio

f ilm

. B

) A z

-sec

tion

of s

ame

pile

up. T

he im

a ge

show

s th

at a

crid

ine

oran

g e o

nly

had

pene

trat

ed in

to th

ede

pth

of 1

0 to

25

µm c

orr e

spon

ding

to a

ppr o

xim

atel

y 10

to 2

0 la

yer s

of c

ells

. Mea

sur e

bar

s ar

e in

µm

. (D

= a

dia

tom

)

BA

54

AB


Fig

ure

18. C

LSM

ima g

es o

f a b

iof il

m o

n a

coup

on o

f sta

inle

ss s

teel

(UN

S S3

1600

) exp

osed

for

5 da

ys in

the

spr a

y w

ater

cir

cuit

of a

pri

ntin

g pa

per

mac

hine

(≈4

5ºC

). A

) A s

ingl

e op

tical

sec

tion

(1 µ

m)

of th

e bi

ofilm

, via

bilit

y st

aine

d w

ith s

yto

9 (l

ive/

gree

n) a

nd p

r opi

dium

iodi

de (

dead

/red

). T

heim

a ge

is ta

k en

from

the

laye

r im

med

iate

to th

e st

eel s

urfa

ce. A

maj

ority

of t

he b

acte

ria

in th

e pe

riph

ery

of th

e bi

ofilm

clu

ster

ar e

red

sta

inin

g de

adba

cter

ia. B

) A s

ingl

e op

tical

sec

tion

(1 µ

m) o

f a b

iof il

m s

tain

ed w

ith e

thid

ium

br o

mid

e (r

ed) a

nd c

arbo

xyla

te-m

odif i

ed fl

uor e

scen

t bea

ds (d

iam

eter

0.02

µm

, gr e

en).

The

im

a ge

was

tak

en a

t th

e di

stan

ce o

f 6

µm t

o st

eel

surf

ace .

The

f igu

r e s

how

s th

at t

hese

flu

ores

cent

bea

ds p

enet

r ate

d to

the

peri

pher

al p

arts

of b

iof il

m c

lust

er o

nly .

Mea

sur e

bar

s ar

e in

µm

.

55

to match the area of the surface of thestainless steel coupons. The emittedfluorescence was recorded from 137 pointson each coupon (diameter 25 mm), andexpressed as relative fluorescence units(rfu). The background autofluorescence ofeach individual coupon was premeasuredbefore placing the coupons in the flow cells.This ensured that minor differences in thefinishing of the individual steel couponswould not interfere with results of thebiofouling measurements.

This novel method permitted the rapidacquisition of quantitative and qualitativeinformation about a living biofilm whilestill attached to its original substratum, aswell as the use of different stains. Thefluorescent methods gave relevantinformation especially at the time when thecoupon surfaces were not yet fully coveredby the biofilm. Scanning fluorometry is lesssuitable for the analysis of thick biofilms.


The biofilms formed on a steel surfaceunder paper machine conditions (high flow,hostile aquatic environment) were thin (≤30 µm, in 7 days). Their ATP content wastoo low to be accurately measured by theprotocols described in the literature. Wedeveloped a novel extraction protocol formeasuring ATP in these biofilms (PaperIII). It involved inserting the steel couponswith the biofilm face downwards into 5 mlof boiling Tris-EDTA with glass beads.During the 5 min of boiling, the glass beadsmechanically disintegrated the cell clusterstightly bound to the steel, releasing the ATPinto the buffer where it could be analysedby the luminometric method (Fig 9 and 11in Paper III). The results showed that it waspossible (detection limit 102 cells cm-2) toextract biofilm ATP from biofouledstainless steel surfaces, and to use the resultas an indicator of biofouling on papermachine steels (Paper III).


3 um

5 um

57Results and discussion

Fig

ure

19. C

LSM

ima g

es o

f a b

iof il

m o

n a

coup

on o

f sta

inle

ss s

teel

(U

NS

S316

00)

espo

sed

for

50 d

ays

in th

e sp

lash

ar e

a of

the

wir

e se

ctio

n of

apr

intin

g pa

per

mac

hine

. Dis

tanc

es o

f the

res

pect

ive

optic

al s

ectio

ns fr

om th

e st

eel s

urfa

ce a

r e in

dica

ted.

The

bio

f ilm

was

via

bilit

y st

aine

d w

ithsy

to 9

(liv

e/gr

een)

and

pr o

pidi

um io

dide

(de

ad/r

ed).

Red

fluo

r esc

enci

ng b

acte

rial

(B

) a g

greg

ates

ar e

vis

ible

am

ong

the

pape

r m

achi

ne f i

nes

and

fibre

s (F

). M

ost p

artic

les

visi

ble

in th

e im

a ges

sho

w a

utof

luor

esce

nce .

Mea

sur e

bar

s ar

e in

µm

.

7 um

9 um

58

5 Summary and conclusions

Biofilm formation in Baltic Sea water during the ennoblement of stainlesssteel started with the attachment of single bacteria. SEM inspection showedSeliberia, Hyphomicrobium and Caulobacter-like bacteria attached to thesteel surface underneath a thin layer of presumably organic substances.Ennoblement of the stainless steels occurred well before the biofilm clustershad fully covered the steel surface. This indicated that the ennoblement ofsteel may be associated with the coexistence of patches with varying accessto oxygen. The patchy, electrically insulating biofilm may form the basis forthe onset of corrosion. We were able to in situ hybridise biofilm-formingbacteria from the steel surface. The majority of the biofilm-forming bacteriain the Baltic Sea belong to the domain bacteria and/or alfa proteobacteria.

The ennoblement of stainless steel reproduced in a laboratory mesocosmwas similar to that occurring in the Baltic Sea. The oligotrophic brackishwater allowed ennoblement to occur in the mesocosm at a flow rate of between10 and 35 mm s-1. At low (0ºC) water temperature, ennoblement in the BalticSea occurred within 50 days, and at 23ºC within 20 to 30 days. A watertemperature of 32ºC reversibly inhibited the onset of ennoblement in thelaboratory mesocosms: ennoblement resumed when the temperature wasdecreased to 26ºC. A two fold increase in the content of nutrients in theBaltic Sea water used as the feed inhibited the onset of ennoblement. Anincrease in the nutrient content also changed the structure of the biofilm.

We showed, using a paper machine splash area simulator, that thiosulphateand oxalate were generated de novo from paper pulp and simulated whitewater. The materials placed in the simulator before the experiment, or thesimulated white water used as the feed, contained no thiosulphate or oxalate.The simulator contained sheets of pulp that were preinoculated with strainsof Bacillus, Desulfovibrio and Desulfotomaculum, as well as an unidentifiedSRB culture from a paper machine environment. Corrosion pits were formedon the stainless steel facing the pulp sheets. The pits occurred in areas wherelarge amounts of thiosulphate and oxalate had accumulated in the pulp sheets.

We constructed flow cells and used them successfully for monitoring biofilmformation in situ in a paper mill. The flow cells were attached to the spraywater circuit of the machine. The biofilm clusters formed under these highflow conditions consisted of bacteria with a uniform shape and size withineach cluster, but with varying shapes and/or sizes between the clusters. Cocci,and short and long rod-shaped bacteria were present in the biofilms, but nofungi or other eukaryotes. Latex beads with a diameter of 0.02 µm penetratedthrough only 2 to 3 layers of cells in the biofilm clusters. In contrast to thebeads, SYTO™ 16, SYTO™9, ethidium bromide and propidium iodide stainspenetrated into the core of the 30 µm-thick biofilm clusters.

Summary and conclusions

I

II

III

IV

59

6 Tiivistelmä

Itämeren vedessä ja paperikoneen prosessivedessä mikrobien tarttuminen ruostumattomaanteräkseen käynnistää biofilmin muodostumisen. Se voi pahimmillaan johtaa putkistojentukkeutumiseen, prosessin teho laskuun, korroosion syntyyn jne.

Itämeren vedessä ensimmäisiä teräkseen tarttuvia mikrobisukuja olivat Seliberia,Hyphomicrobium ja Caulobacter. Näiden lisäksi metallin pinnalle muodostui ohut or-gaanisesta aineesta muodostunut kalvo joka peitti mikrobit. Luonnonvedelle altistetunruostumattoman teräksen lepopotentiaali kasvoi ajan myötä. Tätä tapahtumaa kutsutaanjalontumiseksi. Jalontuminen tapahtui jo silloin kun teräksen pinta oli vasta osittain bio-filmin peittämä. Osittainen peittävyys saattaa aiheuttaa eroja hapen pääsylle metallinpintaan ja näin luoda edellytykset korrosion synnylle. In situ hybridisaatio menetelmällähavaitsimme että suuri osa biofilmimikrobeista oli bakteereja, erityisesti alfa-proteobakteereja.

Teräs jalontui Itämeren vedessä laboratoriossa samalla tavoin kuin meressä. Jalontuminenedellytti riittävän suurta virtausnopeutta (10-35 mm s-1). Jalontumiseen tarvittu aika olipidempi (50 d) kylmässä (0ºC) kuin lämpimässä merivedessä (20-30 d, 23ºC). Merive-den lämpötilan nosto 32ºC:n esti jalontumisen alkamisen mutta lämpötilan laskettuajalontuminen käynnistyi. Myös ravinne pitoisuuksien kaksinkertaistaminen merivedes-sä esti jalontumisen ja muutti biofilmin rakennetta.

Paperikoneen märän pään roiskevyöhykettä simuloivassa laitteessa muodostui tiosulfaattiaja oksalaattia synteettisestä paperikonevedestä ja selluloosasta bakteerien (Bacillus, pa-perikone SRBt, Desulfovibrio ja Desulfotomaculum) läsnäollessa. Tiosulfaattia taioksalaattia ei ollut, tai oli vain pieniä määriä vedessä tai sellussa. Märkien selluloosa-arkkien alla olleeseen ruostumattomaan teräkseen tuli syöpymiä alueille joiden päälläolleesta sellusta löytyi oksalaatia ja tiosulfaattia korkeina pitoisuuksina.

Biofilmin muodostumisen tutkimiseksi paperikoneessa valmistettiin virtauskennot.Kennot kytkettiin viiran suihkuveden vesikiertoon. Voimakas virta (1.8 m s-1) rajoittibiofilmi klustereiden korkeutta. Muodostuneet klusterit koostuivat useimmiten vain yh-denlaisista bakteereista. Bakteerien koko ja muoto vaihtelivat klusterista toiseen.Muodostuneiden biofilmiklustereiden tiiviyttä tutkittiin mittaamalla väriaineidentunkeutumissyvyyttä. Lateksikuulat (halkaisija 0.02 µm) tunkeutuivat vain 2-3 solu-kerroksen syvyyteen. Fluoresoivat väriaineet SYTO™16, SYTO™9, etidiumbromidi japropidiumjodidi tunkeutuivat 30 µm paksun biofilmin läpi.

Väitöskirjatyö osoitti, että biofilmejä voi menestyksellä tutkia suoraan teräspinnoilta,joille mikrobit ovat kasvaneet alkuperäisissa olosuhteissa, kuten tehtaassa tai meressä.Aiempi tietämys biofilmeistä perustui yleensä laboratoriokasvatuksiin lasi- taimuovipinnoilla. Tässä väitöskirjassa kehitetyt menetelmät ja tuotettu tieto biofilmienmikrobirakenteesta ja erilaisten aineiden tunkeutuvuudesta niihin. Tulokset antavat luo-tettavan pohjan prosessiteollisuuden tuotantolaitteiden pesuohjelmien suunnitteluun, eten-kin cleaning-in-place tilanteissa, ja myös mikrobikasvua estävien kemiallisten aineidenja fysikaalisten menetelmien kehittämiseen.

Tiivistelmä

60

7 Acknowledgements

This work was carried out at the Departmentof Applied Chemistry and Microbiology,Division of Microbiology, University ofHelsinki.

The work was supported by EnSTeGraduate School, The Helsinki UniversityFund for Center of Excellence, The Academyof Finland Fund for Center of Excellence,TEKES, Hercules Finland Oy, Metso Oyj,M-real Oyj, Neste Oyj, Outokumpu PolaritOyj and Stora Enso Oyj.

I owe my gratitude to all the following:

My supervisor professor Mirja Salkinoja-Salonen her enthusiasm and endless newideas.

Dos. Anja Klarin-Henricson and Dos.Kirsten Jørgensen for reviewing carefullythis thesis and giving valuable comments.

All my former colleagues in MSS project.The special thanks go to Marko for sharing“lets see how the confocal works”experiments trough the years, to Maria andJoanna keeping up the “normal life“ feelingin our office and to Irina who was reliablesource of microbiological details when everI needed advice. The non-scientificdiscussion, with you people, will carry memuch further than the science never can.

My co-authors Leena Carpén, Laura Raaska,Hanna-Leena Alakomi, Tero Hakkarainen,Assi Weber and Päivi Uutela for theirflexible, reliable and accurate co-operation.

My co-workers during the development of“in the mill”-methods Mats Berg, Karolii-na Karonsuo, Raimo Poutanen, Risto Tal-ja, Pekka Taskinen, Tuovi Valtonen, Juha-ni Vestola, Tarja Vulli and Assi Weber forsharing their expertiece.

Mikrofokus Oy for the flexible use of theirmicroscopes and Simo and Seppo forgiving valuable help with elementalanalysis.

John Derome for language correction andMatti Seppänen for lay out help.

Instrument Center of Faculty of Agricultureand Forestry (specially atk-guys Asko andMara), Laboratory of Electron Microscopyat Helsinki University and Viikki ScienceLibrary for their expertiece service.

Leena (sharing the eye for Lapland’sbeauty), Hannele and Tuula for secretarialhelp. Pauliina for career planning help withcorrect attitude and sharing the joy ofNorjaKurvi. To all staff of MicrobiologyDivision for being friendly and helpful andkeeping things in order and tidy.

Last but definitely not least my warmthanks go to friends whom I wanted to keepout of this work to sustain fresh connectionto real life. You have been sharing the turnsof my life during skiing, paddling, hiking,trekking, cycling and swimming trips.

What I owe to my family mother Liisa,father Antti, brother Jouni and best friendMiska can not be expressed by words.Staying outside from this work they havebeen my strength and they will remain thatafter the persons waiting for they euros andpoints from my thesis have disappeared.

Acknowledgements

Kärkölä, May 2002

61

8 References

Amann RI, L Krumholz and DA Stahl.1990. Fluorescent-oligonucleotideprobing of whole cells for determinative,phylogenetic and environmental studiesin microbiology. J Bacteriol. 172:762-770.

Amann RI, J Stromley, R Devereux, RKey and DA Stahl. 1992. Molecular andmicroscopic identification of sulfate-reducing bacteria in multispecie biofilm.Appl Environ Microbiol. 58:614-623.

Amann RI, W Ludwig and K-H Schleifer.1995. Phylogenetic identification and insitu detection of individual microbialcells without cultivation. Microbiol Rev.59:143-169.

Amann RI, and K-H Schleifer. 2001.Nucleic acid probes and their applicationin environmental microbiology. In:Bergey’s Manual of SystematicBacteriology, Second Edition, Vol 1 (DRBoone, RW Castenholz and GM Garrityeds.), Springer, New York, NY, USA, pp67-82.

Anonymous. 1994. Avesta SheffieldCorrosion Handbook for stainless steels.Avesta Sheffield AB, Sweden.

Anonymous. 1995. Marine biofilm onstainless steels: effects, monitoring andprevention. Report on the Europeanresearch project BIOFILM, contract No.MAS-CT92-0011, pp. 13.

Berthelin J. 1983. Microbial weatheringprocesses. In: Microbial geochemistry.(Krumbein WE, ed), Blackwell ScientificPublications, Oxford, UK. pp.223-262.

Biermann CJ. 1996. Handbook ofpulping and papermaking. AcademicPress, San Diego, CA, USA. Papermanufacture, pp 209-262.

Blanco MA, C Nrgro, I Gaspar and JTijero. 1996. Slime problems in the paperand board industry. Appl MicrobiolBiotechnol. 46:203-208.

Bohlool BB and EL Schmidt. 1980. Theimmunofluorescence approach inmicrobial ecology. Adv Microb Ecol.4:203-241.

Boone DR, RW Castenholz and GMGarrity (eds.). 2001. Bergey’s Manual ofSystematic Bacteriology, Second Edition,Vol 1, Springer, New York, NY, USA, pp721.

Braun-Howland EB, S Danielsen and SANierzwicki-Bauer. 1992. Development ofa rapid method for detecting bacterialcells in situ using 16S rRNA-targetedprobes. Biotechniques. 13:928-933.

Bredholt S, J Maukonen, K Kujanpää, TAlanko, U Olofson, U Husmark, AMSjöberg and G Wirtanen. 1999. Microbialmethods for assessment of cleaning anddisinfection of food-processing surfacescleaned in a low-pressure system. EurFood Res Technol. 209:145-152.

Brown MRW and P Gilbert. 1993.Sensitivity of biofilms to antimicrobialagents. J Appl Bacteriol SymposiumSupplement 74:87S-97S.

Busse H-J, P Kämpfer, ERB Moore, JNuutinen, IV Tsitko, EBM Denner, LVauterin, M Valens, R Rosselló-Mora andMS Salkinoja-Salonen. 2002.Thermomonas haemolytica gen. nov., sp.nov., a g-proteobacterium from kaolineslurry. Int J Systematic and EvolutionaryMicrobiology 52:473-483.

Caldwell D, D Korber and J Lawrence.1992. Confocal laser microscopy andcomputer image analysis in microbialecology. Adv Microb Ecol 12:1-67.

References

62

Carpén L, L Raaska, MS Salkinoja-Salonen and T Hakkarainen. 1995.Microbially induced corrosion ofstainless steel in Baltic Sea water. In:Proceedings of the InternationalConference on Microbially InfluencedCorrosion, May 8th to 10th, 1995, NewOrleans Louisiana USA (P Angell, SWBorenstein, RA Buchanan, SC Dexter,NJE Dowling, BJ Little, CD Dundin, MBMcNeil, DH Pope, RE Tatnail, DC Whiteand HG Ziegenfuss, eds) 1995, NACEInternational, Houston TX, USA, pp. 11/1-11/13.

Carpén L, T Hakkarainen, L Raaska, HAlakomi, MS Salkinoja-Salonen & KMattila. 1997a. Ruostumattomienterästen mikrobikorrosio Itämeren vedes-sä. VTT Raportti VALB225. 27 p &append.

Carpén L, L Raaska, K Mattila, MSSalkinoja-Salonen and T Hakkarainen.1997b. Laboratory Simulation withnatural Bacteria Populations. In: Thierry,D., (ed): Aspects of Microbially InducedCorrosion. European Federation ofCorrosion Publications Number 22, pp113-122.

Carpén L, L Raaska and T Hakkarainen.1999. Effects of biofilm formation on theelectrochemical behaviour of AISI 304SS in board machine environment.Proceedings of CORROSION/99, Paperno. 165. San Antonio, TX, 25-30 April1999. Houston, NACE International, 19p.

Carpén L, L Raaska, K Mattila, MSSalkinoja-Salonen and T Hakkarainen.2000. A novel method to study MIC inboard and paper machine environment.Proceedings of EUROCORR 2000,European Federation of Corrosion, 10.-14.9.2000, London, UK, 10 p.

Carpén L, T Hakkarainen, K Kujanpää, LRaaska, K Mattila, P Uutela & MSSalkinoja-Salonen, 2001a. Biofilmien

vaikutus ruostumattomien terästensyöpymiseen paperikoneympäristöissä.The effect of biofilms on stainless steelsin paper machine environments. Researchreport, in Finnish with English summary.VTT Valmistustekniikka (Espoo),BVAL74-011097. 30 p + 12 append.

Carpén L, L Raaska, K Kujanpää, KMattila, P Uutela, M Salkinoja-Salonen& T Hakkarainen. 2001b. Simulation ofMIC at splash zone areas of the paperindustry. CORROSION/01, Paper No245, Houston TX, 11-16 March 2001.Houston, NACE International, 2001, 16 p

Carpén L, L Raaska, K Kujanpää, KMattila, P Uutela, M Salkinoja-Salonen& T Hakkarainen. 2001c. Laboratorysimulator for splash zone biofouling andcorrosion in paper machine environment.Proceedings of the 10th InternationalSymposium on Corrosion in the Pulp andPaper Industry, Helsinki August 21-24,2001, 12 p

Carpén L, L Raaska, K Kujanpää, KMattila, P Uutela, M Salkinoja-Salonen& T Hakkarainen. 2001d. Simulation ofMIC at splash zone areas of the paperindustry. CORROSION/01, Paper No1245, Houston TX, 11-16 March 2001.Houston, NACE International, HoustonTx, 16 p

Centonze V and J Pawley. 1995. Tutorialon practical confocal microscopy and useof the confocal test specimen. In:Handbook of biological confocalmicroscopy (Pawley JB, ed), PlenumPress, New York, pp 549-569.

Ceri H, ME Olson, C Stremick, RR Read,D Morck and A Buret. 1999. The CalgaryBiofilm Device: New technology forrapid determination of antibioticsusceptibilities of bacterial biofilms. JClinic Microbiol. 37:1771-1776.

Characklis WG. 1991. Biofouling: effectsand control. In: Biofouling and

References

63

biocorrosion in industrial water systems(H-C Flemming and GG Geesey, eds),Springer-Verlag, Berlin, Germany. pp 7-28.

Claus G and R Müller. 1996. Biofilms ina paper mill process water system. InMicrobially influenced corrosion ofmaterials, (E Heitz, H-C Flemming andW Sand, ed), Springer-Verlag, Berlin,Germany. pp 429-437.

Corpe WA and TE Jensen. 1996. Thediversity of bacteria, eukaryotic cells andviruses in an oligotrophic lake. ApplMicrobiol Biotechnol. 46:622-630.

Costerton JW and HM Lappin-Scott.1995. Introduction to microbial biofilms.In Microbial biofilms (HM Lappin-Scottand JW Costerton, eds), Plant andmicrobial biotechnology research series:5, University Press, Cambridge, UK. pp1-11.

Costerton JW, Z Lewandowski, DdeBeer, D Caldwell, D Korber and GJames. 1994. Biofilm, the customizedmicroniche. J Bacteriol. 176:2137-2142.

Costerton JW, Z Lewandowski, DECaldwell, DR Korber and HM Lappin-Scott. 1995. Microbial biofilms. AnnuRev Microbiol. 49:711-745.

Dalton HM, LK Poulsen, P Halasz, MLAngles, AE Goodman and KC Marshall.1994. Substratum-induced morphologicalchanges in a marine bacterium and theirrelevance to biofilm structure. JBacteriol. 176:6900-6906.

Danese PN, LA Pratt and R Kolter. 2000.Exopolysaccharide production is requiredfor development of Escherichia coli K-12biofilm architecture. J Bacteriol.182:3593-3596.

Davey ME and GA O’Toole. 2000.Microbial biofilms: from ecology tomolecular genetics. Microbiology and

molecular biology reviews. 64:847-867.

Davidson CA, CJ Griffith, AC Peters andLM Fielding. 1999. Evaluation of twomethods for monitoring surfacecleanliness- ATP bioluminescence andtraditional hygiene swabbing.Luminescence. 14:33-38.

Davies DG, MR Parsek, JP pearson, BHInglewski, JW Costerton and EP Green-berg. 1998. The involvement of cell-to-cell signalling in the development of abacterial biofilm. Science. 280:295-298.

de Beer D, PD Stoodley and ZLewandowsky. 1997. Liquid flow andmass transport in heterogeneous biofilms.Wat Res. 30:2761-2765.

DeLong EF, GS Wickham and NR Pace.1989. Phylogenetic stains: ribosomalRNA-based probes for the identificationof single microbial cells. Science.243:1360-1363.

Dowling NJE and J Guezennec. 1997.Microbiologically influenced corrosion.In: Manual of EnvironmentalMicrobiology, (CJ Hurst, GR Knudsen,MJ McInerney, LD Stetzenbach and MVWalter, eds), ASM Press, WashingtonD.C., USA, pp 842-855.

Dufrêne YF 2001. Application of atomicforce microscopy to microbial surfaces:from reconstituted cell surface layers toliving cells. Micron. 32:153-165.

Elvers KT, K Leeming, CP Moore andHM Lappin-Scott. 1998. Bacterial-fungalbiofilms in flowing water photo-processing tanks. J Appl Microbiol.84:607-618.

Flemming H-C. 1996. Economical andtechnical overview, In Microbiallyinfluenced corrosion of materials, (EHeitz, H-C Flemming and W Sand, ed),Springer-Verlag, Berlin, Germany. pp 6-14.

References

64

Fletcher M. 1990. Methods for studyingadhesion and attachment to surface, InMethods in microbiology (G Grigorovaand JR Norris, ed), Volume 22,Techniques in Microbial Ecology,Academic Press, London, UK. pp 251-284.

Fuchs BM, G Wallner, W Beisker, ISchwippl, W Ludwig and R Amann.1998. Flow cytometric analysis of the insitu accessibility of Escherichia coli 16SrRNA for fluorescently labelledoligonucleotide probes. Appl. Environ.Microbiol. 64:4973-4982.

Garner A and RC Newman. 1991.Thiosulfate pitting of stainless steels.Corrosion/91, paper no. 186, NACEInternational, Houston, TX, USA.

Geesey GG. 1991. What is corrosion? In:Biofouling and biocorrosion in industrialwater systems (H-C Flemming and GGGeesey, ed), Springer-Verlag, Berlin,Germany. pp 155-164.

Geesey GG. 1993. Biofilm formation. In:A practical manual on microbiologicallyinfluenced corrosion Vol I (Kobrin G,ed), NACE International, Houston, Texas,USA, pp 11-13.

Geesey GG. 2001. Bacterial behavior atsurfaces. Current Opinion inMicrobiology. 4:296-300.

Gharieb MM, J A Sayer and GM Gadd.1998. Solubilization of natural gypsum(CaSO

4•2H

2O) and the formation of

calcium oxalate by Aspergillus niger andSerpula himantioides. Mycol Res.102:825-830.

Gregg CT. 1991. Bioluminescence inclinical microbiology. In Physicalmethods for micro-organisms detection(Nelson WH, ed), CRC Press, BocaRaton, Florida, USA, pp 1-28.

Hakkarainen T, L Carpén, L Raaska, MS

Salkinoja-Salonen and K Mattila. 1996a.Microbially induced corrosion ofstainless steel in natural waters. In: J.Weber & M. Kedro (eds), Interaction ofmicrobial systems with IndustrialMaterials. Helsinki WorkshopProceedings COST 511 (14 to 16 June,1995), ECSC-EC-EAEC, Office forOfficial Publications of the EuropeanCommunities, Brussels-Luxemburg, pp143-146.

Hakkarainen T, L Carpén, L Raaska, KMattila and MS Salkinoja-Salonen1996b. Microbially induced corrosion ofstainless steel in natural waters.Laboratory simulation with naturalbacterial populations. Interaction ofMicrobial Systems with Industrialmaterials, Cost 511 WorkshopProceedings Budapest 30.-31.5.1996,Office for Official Publications of theEuropean Communities, BrusselsLuxembourg, pp 150-155.

Hakkarainen T. 1999. Factors controllingpitting of stainless steel under wetdeposits in splash zones in chlorinedioxide bleaching. Pulp & Paper Can.100:55-58.

Hamilton WA and W Lee. 1995.Biocorrosion. In: Sulfate-reducingbacteria (LL Barton, ed), Biotechnologyhandbooks 8, Plenum Press, New York,USA, pp 243-264.

Haugland RP. 1996. Handbook offluorescent probes and researchchemicals. Molecular probes. Eugene.OR. USA. 679 p.

Heitz E. 1996. Electrochemical andchemical mechanisms. In Microbiallyinfluenced corrosion of materials (EHeitz, H-C Flemming and W Sand, ed),Springer-Verlag, Berlin, Germany. pp 27-38.

Hristova KR, M Mau, D Zheng, RIAminov, RI Mackie, HR Gaskins and L

References

65

Raskin. 2000. Desulfotomaculum genus-and subgenus-specific 16S rRNAhybridization probes for environmentalstudies. Environ Microbiol. 2:143-159.

Ista, LK, VH Pérez-Luna and GP López.1999. Surface-grafted, environmentallysensitive polymers for biofilm release.Appl. Environ. Microbiol. 65:1603-1609.

Kalmbach S, W Manz, B Bendinger andU Szewzyk. 2000. In situ probing revealsAquabacterium commune as awidespread and highly abundant bacterialspecies in drinking water biofilms. WatRes. 34:575-581.

Kepkay PE, SHE Niven and TC Milligan.1993. Low molecular weight andcolloidal DOC production during aphytoplankton bloom. Mar Ecol Prog Ser.100:233-244.

Kolari M, K Mattila, R Mikkola and MSSalkinoja-Salonen. 1998. Communitystructure of biofilms on ennobledstainless steel in Baltic Sea water. JIndustr Microbiol Biotechnol. 21: 261-274.

Kolari M, J Nuutinen and MS Salkinoja-Salonen. 2001. Mechanisms of biofilmformation in paper machine by Bacillusspecies: the role of Deinococcusgeothermalis. J Industr MicrobiolBiotechnol. 27: 343-351.

Kolari M, U Schmidt, E Kuismanen andMS Salkinoja-Salonen. 2002. Firm butslippery attachment of Deinococcusgeothermalis. J Bacteriol. 184:2473-2480.

Kolenbrander PE. 2000. Oral microbialcommunities: biofilms, interactions, andgenetic systems. Annu Rev Microbiol.54:413-437.

Korber DR, JR Lawrence, HM Lappin-Scott and JW Costerton. 1995. Growth ofmicroorganisms on surfaces. In Microbial

biofilms (HM Lappin-Scott and JWCosterton, ed), Plant and microbialbiotechnology research series: 5, Univer-sity Press, Cambridge, UK. pp 15-45.

Korkhaus J, JT Titz and GH Wagner.1996. Pitting corrosion on high-alloyaustenitic steel pipes used for river water.In Microbially influenced corrosion ofmaterials (E Heitz, H-C Flemming and WSand, ed), Springer-Verlag, Berlin,Germany. pp 243-258.

Koskinen, R., T Ali-Vehmas, P Kämpfer,M Laurikkala, I Tsitko, E Kostyal, FAtroshi and MS Salkinoja-Salonen. 2000.Characterization of Sphingomonasisolates from the Finnish and Swedishdrinking water distribution system. JAppl Microbiol. 89:687-696.

Kostyál E. 1998. Removal of chlorinatedorganic matter from wastewater,chlorinated ground, and lake water bynitrifying fluidized-bed biomass.Dissertationes Biocentri ViikkiUniversitatis Helsingiensis 2/1998, Ph.D.thesis, University of Helsinki, Helsinki,Finland

Krumbein WE and D Werner. 1983. Themicrobial silica cycle. In: Microbialgeochemistry. (Krumbein WE, ed),Blackwell Scientific Publications,Oxford, UK. pp.125-157.

Kuehn M, M Hausner, H-J Bungartz, MWagner, PA Wilderer and S Wuertz.1998. Automated Confocal laser scanningmicroscopy and semiautomated imageprocessing for analysis of biofilm. Appl.Environ. Microbiol. 64:4115-4127.

Kuenen G. 1999. Oxidation of inorganiccompounds by chemolithotrophs. In:Biology of the Prokaryotes, (Lengeler,JW, G Drews and HG Schlegel, eds),Georg Thieme Verlag, Stuttgart,Germany. pp.234-260.

References

66

Ladd TI and JW Costerton. 1990.Methods for studying biofilm bacteria. InMethods in microbiology (R Grigorovaand JR Norris, ed), Academic PressLimited, London, UK 22:285-308.

Laitinen T. 1999. Thiosulfate pittingcorrosion of stainless steels in papermachine environment. VTT publications399, Ph.D. thesis, Helsinki TechnicalUniversity, Espoo, Finland.

Lawrence JR, DR Korber, BD Hoyle, JWCosterton and DE Caldwell. 1991.Optical sectioning of microbial biofilms.J Bacteriol. 173:6558-65567.

Lawrence JR, GM Wolfaardt and DRKorber. 1994. Determination of diffusioncoefficients in biofilms by confocal lasermicroscopy. Appl. Environ. Microbiol.60:1166-1173.

Lawrence JR, DR Korber GM Wolfaardtand DE Caldwell. 1997. Analyticalimaging and microscopy techniques. In:Manual of environmental microbiology,(CJ Hurst, GR Knudsen, MJ McInerney,LD Stetzenbach and MV Walter, eds),ASM Press, Washington D.C., USA, pp29-51.

LeChevallier MW. 1991. Biocides andthe current status of biofouling control inwater systems. In: Biofouling andbiocorrosion in industrial water systems(H-C Flemming and GG Geesey, ed),Springer-Verlag, Berlin, Germany. pp114-132.

LeChevallier MW. 1999. Biofilms indrinking water distribution systems:significance and control. In:Indentifyingfuture drinking water contaminants.National Research Council, NationalAcademy Press, Washington, D.C, pp206-219.

Lindberg LE, BR Holmbom, OM Väisä-nen, A M-L Weber and MS Salkinoja-Salonen. 2001a. Sugar composition of

biofilm produced by paper mill bacteria.Appl Microbial Biotechnol. 55:638-643.

Lindberg LE, BR Holmbom, OM Väisä-nen, A M-L Weber and MS Salkinoja-Salonen. 2001b. Degradation of papermill water components in laboratory testswith pure cultures of bacteria.Biodegradation. 12:141-148.

Linhard P. 1996. Failure of chromium-nickel steel in a hydroelectric power plantby manganese-oxidizing bacteria. InMicrobially influenced corrosion ofmaterials, (E Heitz, H-C Flemming andW Sand, ed), Springer-Verlag, Berlin,Germany. pp 221-230.

Lomans BP, HJM Op DEN Camp, A Poland GD Vogels. 1999. Anaerobic versusaerobic degradation of dimethyl sulfideand methanethiol in anoxic freshwatersediments. Appl. Environ. Microbiol.65:438-443.

Lounatmaa K and I Rantala. 1991.Biologinen electronimikroskopia.Yliopistopaino, Helsinki, Finland, 412pp.

Manz W, R Amann, W Ludwig, MWagner and K-H Schleifer. 1992.Phylogenetic oligodeoxynucleotideprobes for the major subclasses ofproteobacteria: problems and solutions.Syst Appl Microbiol. 15:593-600.

Manz W, M Eisenbrecher, TR Neu and USzewzyk. 1998. Abundance and spatialorganization of Gram-negative sulfate-reducing bacteria in activated sludgeinvestigated by in situ probing withspecific 16S rRNA targetedoligonucleotides. FEMS Microbiol Ecol.25:43-61.

Manz W, K Wendt-Potthoff, TR Neu, USzewzyk and JR Lawrence. 1999.Phylogenetic composition, spatialstructure, and dynamics of lotic bacterialbiofilms investigated by fluorescent in

References

67

situ hybridisation and confocal laserscanning microscopy. Microb Ecol.37:225-237.

Marshall KC. 1988. Adhesion and growthof bacteria at surfaces in oligotrophichabitats. Can J Microbiol. 34:503-506.

Marshall KC. 1997. Colonization,adhesion, and biofilms. In: Manual ofEnvironmental Microbiology, (CJ Hurst,GR Knudsen, MJ McInerney, LDStetzenbach and MV Walter, eds), ASMPress, Washington D.C., USA, pp 358-365.

Martínez A, S Torello and R Kolter.1999. Sliding motility in mycobacteria. JBacteriol. 181:7331-7338.

Maukonen J, T Mattila-Sandholm and GWirtanen. 2000. Metabolic indicators forassessing bacterial viability in hygienesampling using cells in suspension andswabbed biofilm. Lebensm.-Wiss. u.-Technol. 33:225-233.

McBride MJ. 2001. Bacterial glidingmotility: multible mechanisms for cellmovement over surface. Annu RevMicrobiol. 55:49-75.

Middelboe M, M Sondergaard, Y Letarteand NH Borch. 1995. Attached and free-living bacteria: production and polymerhydrolysis during a diatom bloom.Microb Ecol. 29:231-248.

Miller MB and BL Bassler. 2001.Quorum sensing in bacteria. Annu RevMicrobiol. 55:165-199.

Morgenroth E and PA Wilderer. 2000.Influence of detachment mechanisms oncompetition in biofilms. Wat Res.34:417-426.

Murga R, TS Forster, E Brown, JMPruckler, BS Fields and RM Donlan.2001. Role of biofilm in the survival ofLegionella pneumophila in a model

potable-water system. Microbiology.147:3121-3126.

Nealson KH and CR Myers. 1992.Microbial reduction of manganese andiron: New approaches to carbon cycling.Appl. Environ. Microbiol. 58:439-443.

Nielsen AT, T Tolker-Nielsen, KB Barkenand S Molin. 2000. Role of commensalrelationships on the spatial structure of asurface-attached microbial consortium.Environ Microbiol. 2:59-68.

Nurmiaho-Lassila E-L, SA Lehtinen, SMarmo and MS Salkinoja-Salonen. 1990.Electron microscopical analysis ofbiological slimes on paper and boardmachines. EMAG-MICRO 89.Proceedings of the Institute of PhysicsConference Series No. 98, Institute ofphysics, Bristol, UK. pp 727-730.

Otero E, JM Bastidas and V López. 1997.Analysis of a premature failure of weldedAISI 316L stainless steel pipes originatedby microbial induced corrosion.Materials and Corrosion. 48:447-454.

O’Toole G, HB Kaplan and R Kolter.2000. Biofilm formation as microbialdevelopment. Annu Rev Microbiol.54:49-79.

Paulus W. 1993. Microbicides for theprotection of materials. Chapman & Hall,London, UK. 496 p

Pawley J. 1995a. Handbook of biologicalconfocal microscopy (Pawley JB, ed),Plenum Press, New York. 632 p

Pawley J. 1995b. Fundamental limits inconfocal microscopy. In: Handbook ofbiological confocal microscopy (PawleyJB, ed), Plenum Press, New York, pp 19-37.

Pedersen K. 1982. Development andrequlation of microbial biofilms in

References

68

flowing water systems. Department ofmarine microbiology, University ofGöteborg, Ph.D. thesis, Göteborg, Swe-den.

Pesonen L, T Norha, I Rinne, I Viitasaloja H Viljamaa 1995. Helsingin ja Espoonmerialueiden velvoitetarkkailu vuosina1987-1994. Helsingin kaupunginympäristökeskus, Moniste 1, Helsinki,Finland, 143 p

Phillips GJ. 2001. Green fluorescentprotein- a bright idea for the study ofbacterial protein localization. FEMSMicrobiol Letters. 204:9-18.

Pirttijärvi TSM, LM Ahonen, LMMaunuksela and MS Salkinoja-Salonen.1998. Bacillus cereus in a whey process.Int J Food Microbiol. 44:31-41.

Pirttijärvi T. 2000. Contaminant aerobicsporeforming bacteria in themanufacturing processes of foodpackaging board and food. DissertationesBiocentri Viikki UniversitatisHelsingiensis 14/2000, Ph.D. thesis,University of Helsinki, Helsinki, Finland

Pirttijärvi TSM, G Wahlström, FARainey, PEJ Saris and MS Salkinoja-Salonen. 2001. Inhibition of bacilli inindustrial starches by nisin. J IndustrMicrobiol Biotechnol. 26:107-114.

Rabus R, M Fukui, H Wilkes and FWiddel. 1996. Degradative capacities and16S rRNA-targeted whole-cellhybridization of sulfate-reducing bacteriain an anaerobic enrichment cultureutilizing alkylbenzenes from crude oil.Appl. Environ. Microbiol. 62: 3605-3613.

Razatos A, Y-L Ong, MM Sharma and GGeorgiou. 1998. Molecular determinantsof bacterial adhesion monitored byatomic force microscopy. Proc Natl AcadSci. USA, 95:11059-11064.

Rice AR, MA Hamilton and AK Camper.2000. Apparent surface associated lagtime in growth of primary biofilm cells.Microb Ecol. 41:8-15.

Rodriques GG, D Phipps, K Ishiguro andHF Ridgway. 1992. Use of a fluorescentredox probe to direct visualization ofactively respiring bacteria. Appl. Environ.Microbiol. 58:1801-1808.

Rossmoore HW (ed.). 1995. Handbookof biocide and preservative use. BlackieAcademic & Professional, Glasgow, UK.424 p.

Salkinoja-Salonen M. 2002.Toisenvaraisten mikrobien aineenvaih-dunta. In: Mikrobiologian perusteita(Salkinoja-Salonen M, ed.). Mikrobiolo-gian julkaisuja 49, Dept. of Appl. Chem.and Microbiology, University of Helsin-ki, Helsinki, Finland. pp 210-274.

Salzburger W. 1996. Paper and additives.In Microbially influenced corrosion ofmaterials (E Heitz, H-C Flemming and WSand, ed), Springer-Verlag, Berlin,Germany. pp 415-427.

Santergoeds CM, LR Damgaard, GHesselink, J Zopfi, P Lens, G Muyzerand D deBeer. 1999. Distribution ofsulfate-reducing and methanogenicbacteria in anaerobic aggregatesdetermined by microsensor andmolecular analyses. Appl. Environ.Microbiol. 65:4618-4629.

Sayer JA, M Kierans and GM Gadd.1997. Solubilisation of some naturallyoccuring metal-bearing minerals,limescale and lead phosphate byAspergillus niger. FEMS MicrobiolLetters. 154:29-35.

Schauder S and BL Bassler. 2001. Thelanguages of bacteria. Genes &Development. 15:1468-1480.

References

69

Schaule G and H-J Flemmen. 1996.Quantification of respiratory activebacteria in water and biofilms with afluorescent redox dye. In Microbiallyinfluenced corrosion of materials (EHeitz, H-C Flemming and W Sand, ed),Springer-Verlag, Berlin, Germany. pp159-165.

Scheuerman TR, AK Camper and MAHamilton. 1998. Effects of substratumtopography on bacterial adhesion. Journalof Colloid and Interface Science. 208:23-33.

Schuster S and GJ Herndl. 1995.Formation and significance of transparentexopolymeric particles in the northernAdriatic Sea. Mar Ecol Prog Ser.124:227-236.

Scotto V and ME Lai. 1998. Theennoblement of stainless steels inseawater: a likely explanations comingfrom the field. Corrosion Science.40:1007-1018.

Shapiro JA. 1998. Thinking aboutbacterial populations as multicellularorganisms. Annu Rev Microbiol. 52:81-104.

Sivonen SJ. 1991.Röntgenmikroanalyysin perusteet. In:Biologinen elektronimikroskopia (KLounatmaa and I Rantala eds.), Yliopisto-paino, Helsinki, Finland, pp 290-312.

Soimajärvi J, M Pursiainen and J Korho-nen. 1978. Sulphate-reducing bacteria inpaper machine waters and in suction rollperforations. European J Appl MicrobiolBiotechnol. 5:87-93.

Stahl DA, R Key, B Flesher and J Smit.1992. The phylogeny of marine andfreswater caulobacters reflects theirhabitat. J Bacteriol. 174:2193-2198.

Starosvetsky D, R Armon, J Yahalom andJ Starosvetsky. 2001. Pitting of carbon

steel caused by iron bacteria. Int Biodetand Biodegrad. 47:79-87.

Stewart PS. 1993. A model biofilmdetachment. Biotech Bioeng. 41:111-117.

Stewart PS, T Grebe, R Srinivasan, C-IChen, FP Yu, D deBeer and GAMcFeters. 1994. Comparision ofrespiratory activity and culturabilityduring monochloramine disinfection ofbinary population biofilms. Appl.Environ. Microbiol. 60:1690-1692.

Stewart PS, AK Camper, SD Handran, C-T Huang and M Warnecke. 1997. Spatialdistribution and coexistence of Klebsiellapneumoniae and Pseudomonasaeruginosa in biofilm. Microb Ecol.33:2-10.

Stoecker JG and DH Pope. 1993. Studyof biological corrosion in hightemperature demineralized water. In: Apractical manual of microbiologicallyinfluenced corrosion (Kobrin G, ed),National Association of CorrosionEngineers, Houston, pp 175-182

Stone JH, MM Gabriel and DG Ahearn.1999. Adherence of Pseudomonasaeruginosa to inanimate polymersincluding biomaterials. J IndustrMicrobiol Biotechnol. 23: 713-717.

Stoodley P, D deBeer and Z Lewandowski.1994. Liquid flow in biofilm system. ApplEnviron Microbiol.. 60:2711-2716.

Stoodley P, Z Lewandowski, JD Boyleand HM Lappin-Scott. 1999. Theformation of migratory ripples in a mixedspecies bacterial biofilm grown inturbulent flow. Environ Microbiol. 1:447-455.

Stoodley P, S Wilson, L Hall-Stoodley, JDBoyle, HM Lappin-Scott and JW Costerton.2001. Growth and detachment of cell clustersfrom mature mixed-species biofilms. Appl.Environ. Microbiol. 67:5608-5613.

References

70

Storgårds E. 2000. Process hygienecontrol in beer production anddispensing. VTT publications 410, Ph.D.thesis, University of Helsinki, Helsinki,Finland

Suominen I, C Spröer, K Lounatmaa, PKäempfer and MS Salkinoja-Salonen.2002. Paenibacillus stellifer, sp nov, acyclodextrin forming species from papermachine. Int J System Evolut Microbiol,submitted.

Szewzyk U, R Szewzyk, W Manz andKH Schleifer. 2000. Microbiologicalsafety of drinking water. Annu RevMicrobiol 54:81-127.

Tatnall RE. 1991. Case histories:biocorrosion. In: Biofouling andbiocorrosion in industrial water systems(H-C Flemming and GG Geesey, eds),Springer-Verlag, Berlin, Germany. pp165-185.

Telegdi J, ZS Kereszter, G Pálinkás, EKálmán and W Sand. 1998. Microbiallyinfluenced corrosion visualized byatomic force microscopy. Appl Phys A.66:S639-S642.

Thorpe PH. 1985. Microbiologicalcorrosion of stainless steel in papermachine. Appita. 40:108-111.

Tolker-Nielsen T and S Molin. 2000a.Spatial organization of microbial biofilmcommunities. Microb Ecol. 40:75-84.

Tolker-Nielsen T, UC Brinch, PC Ragas,JB Andersen, CS Jacobsen and S Molin.2000b. Development and dynamics ofPseudomonas sp. biofilms. J Bacteriol.182:6482-6489.

Vestola J and J Korhonen. 1976.Corrosion and microbiological action insuction rolls. TAPPI. 59:130-133.

Vieira MJ and LF Melo. 1999. Intrinsickinetics of biofilm formed under

turbulent flow and low substrateconcentrations. Bioprocess Engineering.20:369-375.

Vähätalo A. 2000. Role of photochemicalreactions in the biogeochemical cyclingof detrital carbon in aquatic environment.Dissertationes Biocentri ViikkiUniversitatis Helsingiensis 3/2000, Ph.D.thesis, University of Helsinki, Helsinki,Finland

Väisänen O, S Elo, S Marmo and MSalkinoja-Salonen. 1989a. Enzymaticcharacterization of Bacilli from foodpackaging paper and board machines. JIndustrial Microbiology. 4:419-428.

Väisänen O and M Salkinoja-Salonen.1989b. Use of phage typing and fatty acidanalysis for the identification of bacilliisolated from food packaging paper andboard machines. System Appl Microbiol.12:103-111.

Väisänen OM, E-L Nurmiaho-Lassila,SA Marmo and MS Salkinoja-Salonen.1994. Structure and composition ofbiological slime on paper and boardmachines. Appl Environ Microbiol. 60:641-653.

Väisänen OM, A Weber, A Bennasar, FARainey, H-J Busse and MS Salkinoja-Salonen. 1998. Microbial communities ofprinting paper machine. J ApplMicrobiol. 84: 1096-1084.

Watkins Borenstein S. 1994.Microbiologically influenced corrosionhandbook. Woodhead Publishing, Cam-bridge, UK, pp. 288.

Watnick P, KJ Fullner and R Kolter.1999. A role for the mannose-sensitivehemagglutinin in biofilm formation byVibrio cholerae El Tor. J Bacteriol.181:3606-3609.

Watnick P and R Kolter. 2000. Biofilm,city of microbes. J Bacteriol. 182:2675-

References

71

2679.

Weber J and K Knopf. 1996. Generalaspects and selected cases. In Microbiallyinfluenced corrosion of materials (EHeitz, H-C Flemming and W Sand, ed),Springer-Verlag, Berlin, Germany. pp207-220.

Whitehead NA, AML Barnard, H Slater,NJL Simpson and GPC Salmond. 2001.Quorum-sensing in gram-negativebacteria. FEMS, Microbiol Rev. 25:365-404.

Whittaker CJ, CM Klier and PEKolenbrander. 1996. Mechanisms ofadhesion by oral bacteria. Annu RevMicrobiol. 50:513-552

Wimpenny J, W Manz and U Szewzyk.2000. Heterogeneity of biofilms. FEMSMicrobiol Reviews. 24:661-671.

Wingernder J, TR Neu and H-CFlemming (eds.). 1999. Microbialextracellular polymeric substances;characterization, structure and function.Springer-Verlag, Berlin, Germany. p 258.

Wirtanen G. 1995. Biofilm formation andits elimination from food processingequipments. VTT Publications 251, Ph.D.thesis, Helsinki University ofTechnology, Espoo, Finland

Wolfaardt GM, JR Lawrence, RDRobarts, SJ Caldwell and DE Caldwell.1994. Multicellular organization in adegradative biofilm community. ApplEnviron Microbiol. 60:434-446.

Wuertz S, E Müller, R Spaeth, PPfleiderer and H-C Flemming. 2000.Detection of heavy metals in bacterialbiofilms and microbial flocs with thefluorescent complexing agent NewportGreen. J Industr Microbiol Biotechnol.24: 116-123.

Xie H, GS Cook, JW Costerton, G Bruce,

TM Rose and RJ Lamont. 2000.Intergenic communication in dentalplaque biofilms. J Bacteriol. 182:7067-7069.

Yu FP and GA McFeters.1994.Physiological responses of bacteria inbiofilms to disinfection. Appl EnvironMicrobiol. 60:2462-2466.

Zarda B, D Hahn, A Chatzinotas, WSchönhuber, A Neef, RI Amann and JZeyer. 1997. Analysis of bacteriacommunity structure in bulk soil by insitu hybridization. Arch Microbiol.168:185-192.

References

Biofilms on stainless steels exposed to process...

Documents

Transcript of Biofilms on stainless steels exposed to process...