Positively charged biomaterials exert antimicrobial effects on gram-negative bacilli in rats

The development of antimicrobialbiomaterial surfaces

Copyright © 2001, Gottenbos, Bart

The development of antimicrobial biomaterial surfaces

ISBN 90-367-1447-8

Printed by Ponsen & Looijen

Cover design by Aletta van Rheede, Amsterdam

Publication of this thesis was financially supported by:

Faculty of Medical Sciences

Institute of Biomedical Materials Science and Applications

RIJKSUNIVERSITEIT GRONINGEN

The development of antimicrobial

biomaterial surfaces

PROEFSCHRIFT

ter verkrijging van het doctoraat in de

Medische Wetenschappen

aan de Rijksuniversiteit Groningen

op gezag van de

Rector Magnificus, dr. D.F.J. Bosscher,

in het openbaar te verdedigen op

woensdag 13 juni 2001

om 14.15 uur

door

Bart Gottenbos

geboren op 10 september 1973

te Heesch

Promotores : Prof. Dr. Ir. H.J. Busscher

Prof. Dr. P. Nieuwenhuis

Prof. Dr. J. Feijen

Co-promotor : Dr. H.C. van der Mei

Beoordelingscommissie : Prof. Dr. J. Dankert

Prof. Dr. J.E. Degener

Prof. Dr. A.J. Schouten

ISBN 90-367-1447-8

Paranimfen: Albert Poortinga

Gerda Bruinsma

Voor mama

Contents

Chapter 1 Introduction and aims page 9

Chapter 2 Models for studying initial adhesion and surface growth in page 25

biofilm formation on surfaces

Chapter 3 Initial adhesion and surface growth of Staphylococcus page 39

epidermidis and Pseudomonas aeruginosa on biomedical

polymers

Chapter 4 Antimicrobial effects of positively charged surfaces on adhering page 55

Gram-positive and Gram-negative bacteria

Chapter 5 Positively charged biomaterials exert antimicrobial effects on page 69

Gram-negative bacilli in rats

Chapter 6 Late hematogenous infection of subcutaneous implants page 77

in rats

Chapter 7 In vitro and in vivo antimicrobial activity of covalently coupled page 89

quaternary ammonium silane coatings on silicone rubber

Chapter 8 General discussion page 101

Summary page 111

Samenvatting voor niet ingewijden page 115

Nawoord page 119

9

Introduction and aimsBart Gottenbos, Henk J. Busscher, Henny C. van der Mei and Paul Nieuwenhuis

One of the major drawbacks in the use of biomedical materials is the occurrence of

biomaterial-centered infections. After implantation, the host interacts with a biomaterial by

forming a conditioning film on its surface and an immune response towards the foreign

material. When microorganisms can reach the biomaterial surface they can adhere to it.

Adhesion of microorganisms to an implant is mediated by their physico-chemical surface

properties and the properties of the biomaterial surface itself. Subsequent surface growth of

the microorganisms will lead to a mature biofilm and infection, which is difficult to eradicate

by antibiotics. The purpose of this review is to give an overview of the mechanisms involved

in biomaterial-centered infection and the possible strategies to prevent these infections.

Submitted to Journal of Materials Science: Materials in Medicine

1

Chapter 1

10

Introduction

Foreign materials are used more and more in modern medicine after trauma, oncological

surgery or wear to replace, support or restore human body function, for example in hip

prostheses, prosthetic heart valves or catheters. The extended use of these materials, usually

referred to as biomaterials, has some major drawbacks. One of these is the possible

occurrence of biomaterial-centered infections (BCI) [1]. The incidence of this type of

infections varies from 4% for hip prostheses [2] to 100% for urinary tract catheters after 3

weeks use [3] (See Table 1).

Table 1. Incidences of infection of different biomedical implants and devices adapted from Dankert et al. [2] and

arranged according to body site.

Body site Implant or device Incidence (%)

Urinary tract UT catheters 10-20

Percutaneous CV catheters 4-12

Temporary pacemaker 4

Short indwelling catheters 0.5-3

Peritoneal dialysis catheters 3-5

Subcutaneous Cardiac pacemaker 1

Soft tissue Mammary prosthesis 1-7

Intraocular lenses 0.13

Circulatory system Prosthetic heart valve 1.88

Multiple heart valve 3.6

Vascular graft 1.5

Artificial heart* 40

Bones Prosthetic Hip 2.6-4.0

Total knee 3.5-4* From experiments in calves and sheep.

In BCI microorganisms are present in close association with the biomaterial surface forming a

so-called biofilm. BCI can cause severe problems, from disfunctioning of the implanted

device to lethal sepsis of the patient. Furthermore, treatment of BCI is complicated, as

microorganisms in a biofilm are more resistant to antibiotics [4] than their planktonic

counterparts [5]. As a consequence and if possible, the only remedy is removal of the infected

implant at the expense of considerable costs and patient’s suffering. A more convenient way to

Introduction and aims

11

deal with this problem is to prevent the development of an infectious biofilm on the biomaterial

surface. To achieve this, a thorough understanding of how these biofilms develop is necessary.

Host-biomaterials interactions

Conditioning films

The interaction of biomaterials with the body of the host depends on the place where the

implant or device is situated, for example in the oropharyngeal cavity, the urinary tract,

different body tissues or in the circulatory system. When a biomaterial is inserted, first a so-

called conditioning film from organic matter present in the surrounding fluid is deposited on

the biomaterial surface. Depending on the body site the surrounding fluid can be saliva, urine,

tear fluid, tissue fluid, serum or blood, and the conditioning film will mostly consist of

adsorbed proteins, which for serum are mainly albumin, immunoglobulin, fibrinogen and

fibronectin. The composition of the conditioning film depends on the physico-chemical

properties, i.e. chemical composition, hydrophobicity and charge, of the biomaterial surface.

For example, on polyethylene hydrophobicity gradients exposed to blood serum it was found

that at the hydrophobic end less proteins were adsorbed with relatively more fibrinogen, while

on the hydrophilic end more albumin was present [6]. In time, smaller proteins like albumin

are usually replaced by higher molecular weight proteins, like fibrinogen and fibronectin.

With blood contacting biomaterials, also blood cells will adhere to the biomaterial surface.

Especially adhesion of blood platelets to artificial vascular grafts can initiate the blood

coagulation cascade, causing thrombosis, a frequent complication in these applications.

Another unwanted phenomenon at the biomaterial surface is calcification of the implant,

which can decrease the necessary flexibility of, for example, prosthetic heart valves. Ideally

host derived cells will colonize the implanted biomaterials, forming a thin capsule around the

implant.

Immune reactions

The host will also actively interact with the biomaterial surface if it is invasively implanted, as

this is a normal reaction to any foreign body that enters the host. The coagulation cascade and

complement system are activated, leading to formation of a fibrin network and opsonization

of the biomaterial [7, 8]. These processes will attract and activate the innate immune system,

i.e. macrophages and polymorphonuclear cells, leading to inflammation [9,10]. This immune

Chapter 1

12

response can disappear when the wound is healed and the biomaterial is encapsulated.

However, in many cases the host-biomateria ls interface remains in a state of chronic

inflammation, as few metals and plastics are chemically inert in the warm, wet, oxygenated

environment of living tissues, causing the release of inflammatory compounds from the

biomaterial, like corrosion products, plasticizers and monomers [1,11]. Chronic inflammation

impairs host cell growth on the implant [12] and can cause chronic pain.

Pathogenesis of biomaterial-centered infections

The presence of a biomaterial significantly compromises the host to cope with

microorganisms. In a classical study in man it was shown that the presence of a subcutaneous

suture reduced the required inoculum to produce infection with Staphylococcus aureus, an

infamous virulent pathogen, from 106 to only 200 bacteria [13]. Furthermore, the relatively a-

virulent Staphylococcus epidermidis, normally not capable of establishing infection, is the

most common infecting organism in BCI [14].

Inoculation

Probably the most important factor determining the occurrence of BCI is the chance that

microorganisms will reach the biomaterial surface. Biomaterials in contact with the outer part

of the body, for example intravenous catheters, peritoneal dialysis catheters or urinary tract

catheters are readily reached by microorganisms and consequently have a higher incidence of

BCI than fully implanted biomaterials (0.5-100% vs. 0.1-7%) [2]. Microorganisms can reach

a biomaterial implant in several ways at several time points, which determines the properties

of the biomaterial surface they will meet. Airborne microorganisms, which can be present in

the operating theater, can reach the surface as early as before the implantation [15,16] and

interact with a bare substratum surface, not even covered with a conditioning film. Also

during insertion of the biomaterial, microorganisms from the skin can be pushed towards the

implant surface. Furthermore, microorganisms from the skin can contaminate the operation

wound and reach the implant surface through diffusion, active movement or hematogenous

transport. Perioperative contamination is believed to be the most common cause of BCI [17].

Generally, it is also assumed that microorganisms can reach the implant via the

hematogenous route at any time after implantation, causing so-called hematogenous


13

infections. As shown in Table 2 skin infections, surgical or dental interventions, pneumonia,

abscesses or bacteriuria can cause temporal or chronic bacteremia resulting in infections [18].

Table 2. Distant infectious foci of hematogenously infected hip [17] and knee arthroplasties [18].

Distant foci Hip (n=27) (%) Knee (n=72) (%)

Cutaneous region 19 39

Urinary tract 15 19

Respiratory tract 15 14

Oral cavity 30 7

Gastrointestinal tract 4 6

Septic arthritis 0 3

Abdominal abscess 0 1

Unknown 19 11

In addition, microorganisms can translocate from the gastro-intestinal tract to other

body parts [19]. When the microorganisms survive in the bloodstream, they can be

transported to the biomaterial surface, establishing an infection. In this respect it is especially

interesting to note that it has been proposed that macrophages play a role in transporting

microorganisms to the biomaterial surface [20] as some strains are capable to survive within

macrophages [21]. As the biomaterial surface elicits a foreign body reaction in the first weeks

after implantation, and in the case of chronic inflammation also hereafter, macrophages are

specifically attracted to the biomaterial surface thus potentially transporting microorganisms

to the biomaterial implant or device [22]. As hematogenous infection can happen anytime,

biomaterial implants are sometimes called “microbial time bombs”.

The etiology of BCI can provide information about the origin of the infecting

organisms. Table 3 shows the organisms found in examples of studies with vascular grafts,

hip and knee arthroplasties. S. epidermidis and S. aureus, primarily skin inhabitants, are the

predominant infecting organisms [14], followed by Gram-negative bacilli like Escherichia

coli and Pseudomonas aeruginosa, primarily present in the gastro-intestinal and urinary tract,

streptococci from the mouth and pneumococci from the respiratory tract [18,23].

Interestingly, except in the case of airborne microorganisms, the infecting organisms usually

originate from the hosts microflora. There is evidence that the host might be immuno-tolerant

to some of these microorganisms, which has been especially investigated for the intestinal

microflora [24-26]. This might be an overlooked virulence factor for causing BCI. As

Chapter 1

14

immuno-tolerated microorganisms can survive longer in the blood without being attacked by

the hosts immune system, they have a bigger chance to reach an implanted biomaterial.

Table 3. Microbial etiology of vascular graft [23] and hematogenous orthopedic implant [17, 18] infections.

Germ Vascular grafts (%) Orthopedic implants (%)

Abdominal

(n=17)

Inguinal

(n=60)

Popliteal

(n=8)

Hip

(n=27)

Knee

(n=72)

S. aureus 14 40 33 52 52

S. epidermidis 7 13 17 4 4

Streptococci 14 8 25 22 14

Pneumococci n.d. n.d. n.d. 7 6

E. coli 42 9 0 7 11

Proteus species 3 11 0 4 4

other Gr. neg. bacilli 0 8 17 4 4

Other bacteria 10 5 0 0 1

Candida species 3 1 0 0 0

Unknown 7 5 8 0 0

n.d. = not determined. Values for n indicate number of patients

Microbial adhesion

When microorganisms have reached the biomaterial surface, initial microbial adhesion can

occur. Microbial adhesion is mediated by specific interactions between cell surface structures

and specific molecular groups on the substratum surface [14], or when viewed from an

overall, physico-chemical view-point by non-specific interaction forces, including Lifshitz-

Van der Waals forces, electrostatic forces, acid-base interactions and Brownian motion forces

[27]. Specific interactions are in fact non-specific forces acting on highly localized regions of

the interacting surfaces over distances smaller than 5 nm, while non-specific interaction forces

have a long-range character and originate from the entire body of the interacting surfaces, as

is shown in Figure 1 [28]. Upon approach of a surface, organisms will be attracted or repelled

by the surface, depending on the resultant of the different non-specific interaction forces.

Thus the physico-chemical surface properties of the biomaterial, with or without

conditioning film or epithelial cells, and microorganisms play a major role in this process.

The conditioning film on the biomaterial surface (and on the bacterial cell surface) plays an

important role, as it changes the physico-chemical properties of the interacting surfaces. Most

proteins are capable of reducing the adhesion of microorganisms. Albumin is a strong


15

adhesion inhibitor, for unknown reasons, although changes in hydrophobicity and sterical

hindrance are proposed mechanisms [14]. Fibronectin and fibrinogen have been shown to

promote the adhesion of S. aureus and certain S. epidermidis strains, which is mediated by

specific adhesive cell structures directed to these proteins [14,29].

Separationdistance>50 nm

10-20 nm

<5 nm

Substratum

Water

Bacterium

-- -

++ +

Hydrophobic groups

Specificinteractions

Electrostaticinteractions

Short rangeinteractions

-

--

---

-

-

-

-

Figure 1

Figure 1. At separation distances of > 50 nm, only attractive Van der Waals forces occur. At 10 to 20 nm, Van

der Waals and repulsive electrostatic interactions influence adhesion. At < 5 nm, short-range interactions can

occur, irreversibly binding a bacterium to a surface. Adapted from Busscher and Weerkamp [28].

The adhesion mechanism of late hematogenously transported microorganisms is

unclear, as by that time the biomaterial will be covered with host derived cells, which will

decrease the adhesion probability of microorganisms [1]. Hematogenous infections are,

besides with intravascular prostheses, almost exclusively reported with orthopedic implants

[17,18]. Orthopedic implants are usually not completely integrated within host tissue, as they

consist often of metal parts, which are not easily colonized by host tissue cells. The uncovered

metal surface can be colonized readily by microorganisms. Another explanation could be that

the repeated hinging of orthopedic implants, for example in knee prostheses, can cause cell

damage, providing adhesive sites for microorganisms [1].

When the microorganisms have adhered to a biomaterial surface they are protected

against phagocytosis, as the microorganism and biomaterial together are too large to ingest.

Furthermore Zimmerli and coworkers reported that the activity of phagocytes and

polymorphonuclear leukocytes is decreased in the presence of a biomaterial [30,31]. After

adhesion to biomaterials most microorganisms start secreting slime and embed themselves in

Chapter 1

16

a slime layer, the glycocalix, which is an important virulence factor for BCI and which

explains the extraordinary prevalence of slime producing S. epidermidis in BCI [14]. The

glycocalix provides protection against humoral and excreted cellular immune components, as

these can not readily diffuse through the slime layer [4] and once a glycocalix has formed a

BCI with all its complications, including ultimately removal of the implant, seems almost

inevitable. However, to do real damage the adhering microorganisms first have to grow.

Microbial growth

Initial microbial growth, i.e. the growth in the first hours after microbial adhesion, is another

important factor for the outcome of BCI, as biofilm organisms become more resistant to the

immune system while the biofilm matures. Growth of sessile microorganisms has been

mainly studied in vitro [32]. Barton et al. found that growth of sessile P. aeruginosa, S.

epidermidis and E. coli depended on the biomaterial involved. Only for P. aeruginosa they

could find a correlation between physico-chemical properties of the biomaterial surface and

initial growth rates [33]. Also proteins in the condit ioning film can influence the growth rates

of adhering microorganisms. Poelstra et al. found that growth rates of P. aeruginosa

decreased in the presence of pooled immunoglobulin G [34].

Van Loosdrecht et al. [35] concluded that adhesion of bacteria does not directly influence

their metabolism and growth yield. Changes in growth rate due to adhesion of bacteria were

suggested to be mainly the result of changes in nutrient availability. Depending on the amount of

adsorbed nutrients and whether adsorption is easily reversed, growth rates of adhering bacteria

can be decreased or increased with respect to the growth of their planktonic counterparts.

Another mechanism, causing a decrease in the growth rates of E. coli after adhesion was

proposed to be the strong attraction to positively charged biomaterial surfaces [36].

There is little information about microbial growth in vivo. Barth et al. followed the

number of bacteria for 48 hours on subcutaneously implanted polymeric and metal

biomaterials in rabbits, after inoculation at the time of implantation [37]. They found, as is

shown in Figure 2, that the number of non-slime producing S. epidermidis decreased directly

after implantation, probably due to action of the host immune system, while the number of

slime producing S. epidermidis and S. aureus increased in the first 8 to 12 hours to a

maximum whereafter their number decreased again. Here also materials differences played a

role. S. aureus grew faster on the metal, while S. epidermidis grew faster on polymeric

biomaterials. An important difference between in vitro and in vivo experiments is the presence


17

of the immune system. Although unprotected bacteria are eradicated by the immune system, it

has been reported that those microorganisms that do survive accelerate their growth rates

under influence of cytokines excreted by macrophages [38].

A

Time (h)1 4 8 12 20 24 48

CFU

/ di

sc

101

102

103

104

105

106

107

108

B

Time (h)1 4 8 12 20 24 48

CFU

/ di

sc

101

102

103

104

105

106

107

108

C

Time (h)1 4 8 12 20 24 48

CFU

/ di

sc

101

102

103

104

105

106

107

108

Figure 2. The numbers of colony forming units per disk of (a) PMMA, (b) UHMW-PE and (c) titanium

vanadium aluminium alloy, after different in vivo implantation times. SE-360 (black) is a slime producing and

SP-2 (grey) is a non-slime producing strain of S. epidermidis. White bars represent S. aureus. Note the decline in

adhering non-slime producing S. epidermidis in time. Adapted from Barth et al. [37].

Consequences of BCI

The presence of a microbial biofilm on biomaterials impairs the function of the implant or

device and/or worsens the clinical state of the patient. Examples with non-implanted devices

Chapter 1

18

are voice prostheses, which are situated between the trachea and the upper digestive tract, as

is shown in Figure 3, or urinary tract catheters. The action of the voice prosthesis is impaired

by biofilm formation, because microorganisms block the valve mechanism, or cause leakage

of food into the trachea [39]. As a consequence the prosthesis has to be replaced every 4

months on average [40].

Figure 3. Anatomy after total laryngectomy (B). The voice prosthesis (A) is placed in the tracheo-esophageal

shunt (arrow), an area which is heavily inhabited by microorganisms.

Urinary tract catheters rarely escape colonization by microorganisms causing blockage or,

more seriously, bacteriuria [41]. Infections of indwelling catheters, like for example central

venous catheters, often results in bacteremia which can cause sepsis and endocarditis. Totally

implanted prostheses have lower rates of infections, but the consequences are often more

serious. Especially infections of implants in the circulatory system, i.e. prosthetic valves and

vascular grafts, yield a high mortality (70% and 50% respectively [42]). Infection of deep

tissue implants, for example orthopedic implants or mammary prostheses, will usually result

in less serious complications like pain, swelling and loosening of the implant, although

mortalities up to 20% are reported with orthopedic implants [18,43,44]. The clinical signs of

deep implant infections are reported to appear up to a year after microbial seeding [45].

Apparently the infectious biofilm can stay silent for a long period, and probably a significant

part of the infections is never recognized. Recent investigations at our laboratory revealed that

BA

esophagealretentionflange

valve shaft

openend

trachealretentionflange

valve

A B

esophagus

trachea


19

the so-called aseptic loosening, i.e. loosening of an orthopedic implant without

microorganisms present, is often diagnosed falsely as a-septic. As standard microbiological

techniques are used to test for the presence of infectious microorganisms, slow growing

biofilm organisms often remain undetected. Similarly, it has been reported that 5 of 28

removed scleral explants were covered with a biofilm, while clinical signs of infection were

only present in 1 out of these 5 cases [46]. Thus the incidence of problems associated with

BCI is possibly higher than generally assumed.

Treatment and prevention of BCI

Treatment

Treatment of an established BCI is difficult, as the minimal inhibitory concentration (MIC) of

antimicrobial agents, necessary to kill the microorganisms, is significantly higher for

microorganisms in a biofilm than for planktonic ones [4,5]. As antibiotics have little effect on

BCI, the standard procedure for infected orthopedic prostheses is the removal of the implant

and implantation of an antibiotic releasing device at the implant site. A new prosthesis is

inserted when the implant site is free of microorganisms, usually six months later. For many

implants, especially those in the circulatory system, removal of the implant is dangerous, and

a high mortality is associated with these infections. Much research has been done to make

biofilm bacteria more susceptible to antibiotics. Ultrasonic treatment of the infecting biofilm,

for example, has been shown to enhance the action of antibiotics towards these biofilm

bacteria [47]. Also application of an electrical field yields enhanced effects of antibiotic

treatment, the so-called bioelectric effect [48]. The application of these techniques in patients

could facilitate the treatment of BCI with antibiotics in the future.

Prevention

Surgeons take considerable effort in preventing the contamination of implants with

microorganisms during implantation. Although application of prophylactic antibiotics and

better operation hygiene has reduced the incidence of BCI the last four decades, still a

significant number of patients suffer from these infections [2].

Different strategies seem useful to prevent BCI. In general it is aimed to reduce the

attractive force between bacteria and biomaterial surface by optimizing the physico-chemical

surface properties of the biomaterial. Bacterial adhesion is low, for example, on extremely

Chapter 1

20

hydrophobic surfaces [49,50], while also more negatively charged biomaterials attract less

bacteria [51]. Albumin and heparin coatings have shown to decrease the adhesiveness of

biomaterials [52].

However, microorganisms always seem to be able to adhere to some extent to solid

materials. Moreover, when proteins are present they can cover an anti-adhesive biomaterial,

and be anchors for microorganisms to adhere to. Another approach to prevent biofilm

formation is to prevent the growth of adhering microorganisms. This can be achieved by

application of antimicrobial agents near the biomaterial surface. One way to do this is the

design of antibiotic releasing biomaterials. Examples are gentamicin-loaded bone cement and

silver-loaded catheters [53,54]. A disadvantage of these applications is that they usually only

work for a few days to weeks, as the amount of antibiotic that is actually released is extremely

limited and does not exceed 15% of the total amount incorporated [55]. A more dangerous

problem with antibiotic releasing materials and the low dose actually released is the

development of antibiotic resistant microbial strains [56]. A better approach would be to

couple an antimicrobial agent covalently onto the biomaterial surface, while maintaining its

activity. As in this approach the antimicrobial agent can only reach the outside of the

microbial cells, it can only be employed with antibiotics working at the level of the cell wall

or membrane. Polymers with incorporated quaternary ammonium groups have shown such

antimicrobial activity in vitro [57,58], thus these compounds might have the required

properties.

Aims of this thesis

The aim of the research described in this thesis was to analyze the biomaterial surface

properties that influence bacterial growth. Subsequently, a second aim was to develop

biomaterials with antimicrobial surface properties, i.e. on which bacteria can not grow,

without leaking of antimicrobial agents from the surface. Finally, as a third aim, a model was

developed to determine the antimicrobial properties of the newly designed biomaterial

surfaces in vivo.


21

References

1. Gristina, A.G. (1987). Biomaterial-centered infection: microbial adhesion versus tissue integration.

Science 237, 1588-95.

2. Dankert, J., Hogt, A.H. and Feijen, J. (1986). Biomedical polymers: Bacterial adhesion, colonization and

infection. CRC Critical Reviews in Biocompatibility 2, 219-301.

3. Denstedt, J.D., Wollin, T.A. and Reid, G. (1998). Biomaterials used in urology: current issues of

biocompatibility, infection, and encrustation. Journal of Endourology 12, 493-500.

4. Costerton, J.W., Stewart, P.S. and Greenberg, E.P. (1999). Bacterial biofilms: a common cause of

persistent infections. Science 284, 1318-22.

5. Gilbert, P., Das, J. and Foley, I. (1997). Biofilm susceptibility to antimicrobials. Advances in Dental

Research 11, 160-7.

6. Spijker, H.T., Bos, R., Van Oeveren, W., De Vries, J. and Busscher, H.J. (1999). Protein adsorption on

gradient surfaces on polyethylene prepared in a shielded gas plasma. Colloids and Surfaces B:

Biointerfaces 15, 89-97.

7. Tang, L. and Eaton, J.W. (1995). Inflammatory responses to biomaterials. American Journal of Clinical

Pathology 103, 466-71.

8. Remes, A. and Williams, D.F. (1992). Immune response in biocompatibility. Biomaterials 13, 731-43.

9. Mollnes, T.E. (1998). Complement and biocompatibility. Vox Sanguinis, Supplement 2 74, 303-7.

10. Anderson, J.M. (1988). Inflammatory response to implants. ASAIO Transactions 34, 101-7.

11. Dougherty, S.H. and Simmons, R.L. (1982). Infections in bionic man: the pathobiology of infections in

prosthetic devices--Part I. Current Problems in Surgery 19, 217-64.

12. Jackson, J.H. and Cochrane, C.G. (1988). Leukocyte-induced tissue injury. Hematology Oncology

Clinics of North America 2, 317-34.

13. Elek, S.D. and Conen, P.E. (1957). The virulence of Staphylococcus pyogenes for man. A study of the

problems of wound infection. British Journal of Experimental Pathology 38, 573-86.

14. Christensen, G.D., Baddour, L.M., Hasty, D.L., Lowrance, J.H. and Simpson, W.A. (1989). Microbial

and foreign body factors in the pathogenesis of medical device infections. In Infections associated with

indwelling medical devices. Edited by Bisno, A.L. and Waldvogel, F.A. (American Society of

Microbiology, Washington DC) p. 27-59.

15. Lidwell, O.M., Lowbury, E.J., Whyte, W., Blowers, R., Stanley, S.J. and Lowe, D. (1982). Effect of

ultraclean air in operating rooms on deep sepsis in the joint after total hip or knee replacement: a

randomised study. British Medical Journal Clinical Research Edition 285, 10-4.

16. Charnley, J. (1972). Postoperative infection after total hip replacement with special reference to air

contamination in the operating room. Clinical Orthopaedics and Related Research 87, 167-87.

17. Ahlberg, A., Carlsson, A.S. and Lindberg, L. (1978). Hematogenous infection in total joint replacement.

Clinical Orthopaedics and Related Research 137, 69-75.

18. Bengtson, S., Blomgren, G., Knutson, K., Wigren, A. and Lidgren, L. (1987). Hematogenous infection

after knee arthroplasty. Acta Orthopaedica Scandinavica 58, 529-34.

Chapter 1

22

19. Van Leeuwen, P.A., Boermeester, M.A., Houdijk, A.P., Ferwerda, C.C., Cuesta, M.A., Meyer, S. and

Wesdorp, R.I. (1994). Clinical significance of translocation. Gut 35, S28-S34.

20. Mora, E.M., Cardona, M.A. and Simmons, R.L. (1991). Enteric bacteria and ingested inert particles

translocate to intraperitoneal prosthetic materials. Archives of Surgery 126, 157-63.

21. Wells, C.L., Maddaus, M.A. and Simmons, R.L. (1987). Role of the macrophage in the translocation of

intestinal bacteria. Archives of Surgery 122, 48-53.

22. Guo, W., Andersson, R., Ljungh, A., Wang, X.D. and Bengmark, S. (1993). Enteric bacterial

translocation after intraperitoneal implantation of rubber drain pieces. Scandinavian Journal of

Gastroenterolology 28, 393-400.

23. Printzen, G. (1996). Relevance, pathogenicity and virulence of microorganisms in implant related

infections. Injury, Supplement 3 27, SC9-15.

24. Duchmann, R., Schmitt, E., Knolle, P., Meyer zum Buschenfelde, K.H. and Neurath, M. (1996).

Tolerance towards resident intestinal flora in mice is abrogated in experimental colitis and restored by

treatment with interleukin-10 or antibodies to interleukin-12. European Journal of Immunology 26, 934-8.

25. Foo, M.C. and Lee, A. (1972). Immunological response of mice to members of the autochthonous

intestinal microflora. Infection and Immunity 6, 525-32.

26. Berg, R.D. and Savage, D.C. (1975). Immune responses of specific pathogen-free and gnotobiotic mice to

antigens of indigenous and nonindigenous microorganisms. Infection and Immunity 11, 320-9.

27. Van Oss, C.J. (1991). The forces involved in bioadhesion to flat surfaces and particles - their

determination and relative roles. Biofouling 4, 25-35.

28. Busscher, H.J. and Weerkamp, A.H. (1987). Specific and non-specific interactions in bacterial adhesion

to solid substrata. FEMS Microbiology Reviews 46, 165-73.

29. An, Y.H. and Friedman, R.J. (1998). Concise review of mechanisms of bacterial adhesion to biomaterial

surfaces. Journal of Biomedical Materials Research 43, 338-48.

30. Zimmerli, W., Lew, P.D. and Waldvogel, F.A. (1984). Pathogenesis of foreign body infection. Evidence

for a local granulocyte defect. Journal of Clinical Investigations 73, 1191-200.

31. Zimmerli, W., Waldvogel, F.A., Vaudaux, P. and Nydegger, U.E. (1982). Pathogenesis of foreign body

infection: description and characteristics of an animal model. Journal of Infectious Diseases 146, 487-97.

32. Geesey, G.G. and White, D.C. (1990). Determination of bacterial growth and activity at solid-liquid

interfaces. Annual Review of Microbiology 44, 579-602.

33. Barton, A.J., Sagers, R.D. and Pitt, W.G. (1996). Measurement of bacterial growth rates on polymers.

Journal of Biomedical Materials Research 32, 271-8.

34. Poelstra, K.A., Van der Mei, H.C., Gottenbos, B., Grainger, D.W., Van Horn, J.R. and Busscher, H.J.

(2000). Pooled human immunoglobulins reduce adhesion of Pseudomonas aeruginosa in a parallel plate

flow chamber. Journal of Biomedical Materials Research 50, 224-32.

35. Van Loosdrecht, M.C.M., Lyklema, J., Norde, W. and Zehnder, A.J.B. (1990). Influence of interfaces

on microbial activity. Microbiological Reviews 54, 75-87.

36. Harkes, G., Dankert, J. and Feijen, J. (1992). Growth of uropathogenic Escherichia coli strains at solid

surfaces. Journal of Biomaterial Science: Polymer Edition 3, 403-18.


23

37. Barth, E., Myrvik, Q.M., Wagner, W. and Gristina, A.G. (1989). In vitro and in vivo comparative

colonization of Staphylococcus aureus and Staphylococcus epidermidis on orthopaedic implant materials.

Biomaterials 10, 325-8.

38. Van Wachem, P.B., Van Luyn, M.J.A., De Wit, A.W., Raatjes, D., Hendriks, M., Verhoeven,

M.L.P.M. and Cahalan, L. (1997). Tissue reactions to bacteria-inoculated rat lead samples. I. Effect of

local gentamicin release through vicinal sponge or solution-dipping. Journal of Biomedical Materials

Research 35, 217-32.

39. Mahieu, H.F., Saene, H.F., Rosingh, H.J. and Schutte, H.K. (1986). Candida vegetations on silicone

voice prosthesis. Archives of Otolaryngology 112, 321-5.

40. Van den Hoogen, F.A., Oudes, M.J., Hombergen, G., Nijdam, H.F. and Manni, J.J. (1996). The

Groningen, Nijdam and provox voice prostheses: a prospective clinical comparison based on 845

replacements. Acta Otolaryngologica 116, 119-24.

41. Nickel, J.C., Costerton, J.W., McLean, R.J. and Olson, M. (1994). Bacterial biofilms: influence on the

pathogenesis, diagnosis and treatment of urinary tract infections. Journal of Antimicrobial Chemotherapy,

supplement A 33, 31-41.

42. Mayer, K.H. and Schoenbaum, S.C. (1982). Evaluation and management of prosthetic valve endocarditis.

Progress in Cardiovascular Diseases 25, 43-54.

43. Hunter, G. and Dandy, D. (1977). The natural history of the patient with an infected total hip

replacement. The Journal of Bone and Joint Surgery, British volume 59, 293-7.

44. Fitzgerald, R.H. and Jones, D.R. (1985). Hip implant infection. Treatment with resection arthroplasty and

late total hip arthroplasty. American Journal of Medicine 78, 225-8.

45. Maniloff, G., Greenwald, R., Laskin, R. and Singer, C. (1987). Delayed postbacteremic prosthetic joint

infection. Clinical Orthopaedics and Related Research 223, 194-7.

46. Asaria, R.H., Downie, J.A., McLauglin-Borlace, L., Morlet, N., Munro, P. and Charteris, D.G. (1999).

Biofilm on scleral explants with and without clinical infection. Retina 19, 447-50.

47. Rediske, A.M., Roeder, B.L., Brown, M.K., Nelson, J.L., Robison, R.L., Draper, D.O., Schaalje, G.B.,

Robison, R.A. and Pitt, W.G. (1999). Ultrasonic enhancement of antibiotic action on Escherichia coli

biofilms: an in vivo model. Antimicrobial Agents and Chemotherapy 43, 1211-4.

48. Costerton, J.W., Ellis, B., Lam, K., Johnson, F. and Khoury, A.E. (1994). Mechanism of electrical

enhancement of efficacy of antibiotics in killing biofilm bacteria. Antimicrobial Agents and Chemotherapy

38, 2803-9.

49. Everaert, E.P., Mahieu, H.F., Van de Belt-Gritter, B., Peeters, A.J., Verkerke, G.J., Van der Mei,

H.C. and Busscher, H.J. (1999). Biofilm formation in vivo on perfluoro-alkylsiloxane-modified voice

prostheses. Archives of Otolaryngology: Head and Neck Surgery 125, 1329-32.

50. Tsibouklis, J., Stone, M., Thorpe, A.A., Graham, P., Peters, V., Heerlien, R., Smith, J.R., Green, K.L.

and Nevell, T.G. (1999). Preventing bacterial adhesion onto surfaces: the low-surface-energy approach.

Biomaterials 20, 1229-35.

51. Hogt, A.H., Dankert, J. and Feijen, J. (1986). Adhesion of coagulase-negative staphylococci to

methacrylate polymers and copolymers. Journal of Biomedical Materials Research 20, 533-45.

Chapter 1

24

52. Keogh, J.R. and Eaton, J.W. (1994). Albumin binding surfaces for biomaterials. Journal of Laboratory

and Clinical Medicine 124, 537-45.

53. Carlsson, A.S., Josefsson, G. and Lindberg, L. (1978). Revision with gentamicin-impregnated cement for

deep infections in total hip arthroplasties. The Journal of Bone and Joint Surgery 60, 1059-64.

54. Akiyama, H. and Okamoto, S. (1979). Prophylaxis of indwelling urethral catheter infection: clinical

experience with a modified Foley catheter and drainage system. Journal of Urology 121, 40-2.

55. Van de Belt, H., Neut, D., Uges, D.R.A., Schenk, W., Van Horn, J.R., Van der Mei, H.C. and

Busscher, H.J. (2000). Surface roughness, porosity and wettability of gentamicin-loaded bone cements and

their antibiotic release. Biomaterials 21, 1981-7.

56. Van de Belt, H., Neut, D., Van Horn, J., Van der Mei, H.C., Schenk, W. and Busscher, H.J. (1999).

...or not to treat?. Nature Medicine 5, 358-9.

57. Flemming, R.G., Capelli, C.C., Cooper, S.L. and Proctor, R.A. (2000). Bacterial colonization of

functionalized polyurethanes. Biomaterials 21, 273-81.

58. Kenawy, E., Abdel-Hay, F.I., El-Shanshoury, A.R. and El-Newehy, M.H. (1998). Biologically active

polymers: synthesis and antimicrobial activity of modified glycidyl methacrylate polymers having a

quaternary ammonium and phosphonium groups. Journal of Controlled Release 50, 145-52.

25

Models for studying initial adhesion and surface

growth in biofilm formation on surfacesBart Gottenbos, Henny C. van der Mei and Henk J. Busscher

Microbial biofilms cause various problems in industry, waterworks, dentistry and medicine.

In this chapter a parallel plate flow chamber system is described, with which different

processes in bacterial biofilm formation, i.e. conditioning film formation, initial bacterial

adhesion, bacterial surface growth and bacterial detachment can be modeled and monitored in

situ. Examples are given of studies concerning the influence of a plasma conditioning film on

initial adhesion, the influence of biomaterial surface properties on surface growth and the

influence of surface active substances on detachment of biofilm bacteria.

Reproduced with permission of Academic Press from

Methods in Enzymology 310 (1999) 523-33

2

Chapter 2

26

Introduction

Biofilms can be considered as micro-ecosystems in which different microbial strains and

species efficiently cooperate in order to protect themselves against environmental stresses,

and to facilitate more efficient nutrient uptake. Most often, biofilms are unwanted, and related

to diverse problems as microbially induced corrosion of oil-rigs and pipelines [1], food and

drinking water contamination [2,3], dental caries and periodontal diseases [4] and a variety of

biomaterial-centered infections in man [5]. Biomaterial-centered infections in man are

especially troublesome, since biofilm organisms are protected against the host immune system

and cannot be easily eradicated with antibiotics. Consequently, infection of a biomaterial

implant will usually result in reoperation, osteomyelitis, amputation or death [6]. Not all

biofilms are unwanted, however, and in sewage treatment, biofilms are needed for efficient

degradation of xenobiotics [7], while lactobacillus biofilms form part of the normal

indigenous microflora in man and their maintenance is essential in the prevention of disease

[8].

Mechanism of biofilm formation

The formation of a biofilm in an aqueous environment is generally pictured to proceed in the

following sequence [9] (see Figure 1):

1. When organic matter is present, a conditioning film of adsorbed components is formed on

the surface prior to the arrival of the first organisms

2. Microorganisms are transported to the surface through diffusion, convection,

sedimentation or active movement

3. Initial microbial adhesion occurs

4. Attachment of adhering microorganisms is strengthened through exopolymer production

and unfolding of cell surface structures

5. Surface growth of attached microorganisms and continued secretion of exopolymers

6. Localized detachment of biofilm organisms caused by occasionally high fluid shear or

other detachment forces operative

Localized detachment of biofilm organisms starts after initial adhesion, although adhesion of

individual microorganisms is frequently considered irreversible (whether justified or not), and

increases with time as it is related to the number of microorganisms present in the biofilm

[10].

Models for studying initial adhesion

27

Figure 1. Schematic, sequential presentation of the steps in biofilm formation.

Detachment of parts of a biofilm can occur by failure inside the bulk of the biofilm, or by

failure in the so-called linking film, involving either detachment of the initially adhering

organisms, cohesive failure in conditioning film or interfacial rupture. Furthermore, as the

Chapter 2

28

number of biofilm organisms increases, growth rates will decrease due to nutrient and oxygen

limitations and accumulation of organic acids, eventually leading to a stationary biofilm

thickness, where adhesion and growth counterbalance detachment.

Prevention, control and eradication of biofilms

An obvious approach in the prevention of biofilm formation is the prevention of initial

microbial adhesion. Microbial adhesion is mediated by specific interactions between cell

surface structures and specific molecular groups on the substratum surface [11], or when

viewed from an overall, physico-chemical view-point by non-specific interaction forces,

including Lifshitz-Van der Waals forces, electrostatic forces, acid-base interactions and

Brownian motion forces [12]. Specific interactions are in fact non-specific forces acting on

highly localized regions of the interacting surfaces over distances smaller than 5 nm, while

non-specific interaction forces have a long range character and originate from the entire body

of the interacting surfaces. Upon approach of a surface, organisms will be attracted or repelled

by the surface, depending on the resultant of the different non-specific interaction forces.

Lifshitz-Van der Waals forces and Brownian motion usually promote adhesion, while

electrostatic interactions can be either attractive or repulsive. Most organisms are negatively

charged [5] and consequently a negatively charged substratum exerts a repulsive electrostatic

force on the organisms. Control of the charge and hydrophobic properties of substratum

surfaces is likewise a pathway to influence biofilm interaction with a substratum surface.

Another pathway to influence biofilm formation is through inhibition of growth. A

possible approach is the design of (antibiotic)slow-release materials mediating direct kill upon

contact [13], but such approaches are always temporarily and bear the risk of inducing

resistant strains. Alternatively, also physico-chemical surface properties affect growth and it

has been found that Escherichia coli growth is inhibited on positively charged surfaces

through strong attachment [14].

To control beneficial biofilms, conditions have to be adapted to obtain the optimal

equilibrium of growth and detachment of biofilm organisms. Growth rate can be controlled by

nutrient conditions, reactor design, oxygen household, removal of metabolic end-products and

temperature. Detachment can be influenced by substratum surface properties, fluid flow rate

[15], mechanical stress and surface active substances [16].

Eradication of unwanted biofilms with disinfectants or antibiotics is hampered,

because of the resistance of biofilm organisms against anti-microbial penetration through the


29

biofilm [17]. Total removal of biofilms is only possible, when biofilms are directly accessible,

as e.g. on exterior parts of the human body. Dental biofilms can be removed by mechanical

cleansing, in combination with the use of surface active substances, like in dentifrices and

mouth washes. Bacteria adhering on contact lenses can be removed by rubbing the lenses

between the fingers in combination with cleansing solutions. The strength of biofilm adhesion

to a substratum surface, i.e. the ease with which it can be removed, is greatly dependent on

the strength with which the initially adhering organisms bind the substratum surface and

cohesiveness of the conditioning film [18], which makes initial microbial adhesion and

surface growth an important issue of research.

This chapter summarizes the use of a parallel plate flow chamber model to study

initial microbial adhesion to surfaces and extends the use of flow chamber devices and data

analysis to include surface growth of the initially adhering organisms.

Figure 2. Detailed view of the parallel plate flow chamber.

Experimental design

The parallel plate flow chamber and image analysis used in our laboratory is accurately

described in an earlier volume of this series [19], but will be briefly repeated for

completeness. Figure 2 shows the chamber (external dimensions 16 x 8 x 2 cm, length by

width by height) which is made of nickel-coated brass to allow sterilization. The internal

dimensions of the chamber are 7.6 x 3.8 x 0.06 cm, although the height can be varied by using

different spacers. Microbial adhesion, surface growth and detachment is directly observed

Chapter 2

30

usually on the bottom plate using phase-contrast microscopy, thus biofilm formation can be

followed in situ without any additional shear forces acting on the deposited microorganisms.

The bottom plate can be of various materials but can be conveniently made of poly(methyl

methacrylate) (PMMA) with a grove in the middle in which a substratum material of interest

can be fixed. The top plate is made of glass. The application of phase-contrast microscopy

requires transparent substratum materials. Non-transparent, reflective substrata can also be

used when an ordinary metallurgical microscope, based on incident reflected light, is

available [20]. To allow the detection of microorganisms on opaque materials incandescent

dark-field illumination is directed under a low angle onto the bottom plate of the channel by

means of a lens supported on a slide [21]. Furthermore, it is possible to use fluorescence

microscopy, provided it is ascertained that the required fluorescent dyes do not affect the

adhesive properties of the microorganisms.

The microscope images can be recorded using a CCD camera and processed by an

image analyzer (TEA, Image-Manager, Difa Breda, the Netherlands) in combination with

dedicated image analysis programs [22].

Flow is created by a roller pump and can be pulse-free or controlled pulsatile [23].

Optional is temperature control using heating elements mounted on the sides of the chamber.

By means of a valve system it is possible to connect flasks containing, for example, buffer,

reconstituted human whole saliva, bacterial suspension, growth medium or detergent

solutions, with the flow chamber without passing an air-liquid interface over the adsorbed

conditioning film and/or adhering organisms [24].

Initial microbial adhesion

Initial microbial adhesion experiments are typically done in buffer solutions without any

additional nutrients, to avoid complications caused by conditioning films or growth of

microorganisms during deposition. Before each experiment all air bubbles are removed from

the tubing and flow chamber and the buffer solution is perfused through the system for a

predetermined time. Subsequently, flow is switched to a suspension of microorganisms in

buffer. During deposition, images of the bottom plate are recorded and the organisms present

on the surface are counted using the image analysis program. From the number of

microorganisms plotted versus time the initial deposition rate (j0) is determined and through

an iterative procedures the number of bacteria adhering at a stationary end-point time (n∞) is

found.


31

As natural biofilms are not formed in plain buffer, it is relevant to study adhesion also

in the fluids present in nature. For example in studies on biofilm formation on teeth [25],

urinary catheters [26], contact lenses [16] or body implants [27], buffer can be replaced by

saliva, urine, tear fluid or blood plasma. Here initial adhesion can be influenced by the

composition of the conditioning film. In blood serum, for example, albumin reduces

staphylococcal adhesion [27], while fibronectin can promote adhesion of certain

staphylococcal strains [28,29]. The contribution of surface growth to the number of attached

bacteria cannot easily be separated from that of adhesion, which complicates these types of

experiments.

Surface growth

The use of the parallel plate flow chamber can be extended to study surface growth of sessile

microorganisms [30,31]. For this purpose, the entire system is sterilized at 120°C, except for

the PMMA bottom plate and the substratum material of interest, which are sterilized by 70%

ethanol. During the experiment the flow chamber can be heated to 37°C to get more relevant

results for surface growth on biomaterials. The experiment starts with initial adhesion of the

microorganisms during a short duration of time, as microorganisms tend to loose their

viability in buffer. This can be avoided, however, by supplementing the buffer with growth

medium. For Pseudomonas aeruginosa, for example, it was found that only 2% of the

adherent bacteria was metabolically active after 4 h deposition in phosphate buffered saline

(PBS), while in a minimal (2%) growth medium 67% appeared metabolically active [32]. In

the above case, addition of a minor amount of growth medium did not seriously complicate

interpretation of the results.

After microorganisms are seeded on the substratum surface, the flow chamber is

washed with adhesion buffer at the same flow rate as during deposition, to remove the

planktonic and loosely adhering organisms. Subsequently, flow is switched to growth medium

at the same flow rate and continued for a predetermined time. Images are recorded during

surface growth, from which the number of organisms present is determined. With appropriate

analysis, it is possible to follow individual microorganisms and determine their generation

time.

Chapter 2

32

Eradication of biofilms

After microorganisms are deposited or grown in the flow chamber, various effects of

environmental stress on the biofilm can be studied. To study the influence of high shear stress

on the adhering organisms, we usually pass an air bubble over the substratum surface [33].

The passing of a liquid-air interface results in a detachment force on the adhering organisms

of around 10-7 N [19], which is much higher than the shear resulting from the flowing liquid.

The number of organisms present before, and after the passage of the liquid-air interface can

be determined as a measure of the strength of microbial adhesion.

To determine detergent-stimulated detachment of biofilm organisms, a surfactant

solution can be led over the adhering microorganisms. Images are recorded before, during and

after the treatment, and the microorganisms are enumerated. In this way, the activity of, for

example, mouth washes and contact lens cleansing solutions can be evaluated.

Examples of the use of parallel plate flow chamber for studying biofilm

formation and eradication

Influence of a blood plasma conditioning film on initial staphylococcal adhesion

For six staphylococcal strains, staphylococcal adhesion was studied on silicone rubber with

and without pre-adsorbed plasma proteins [27]. First, when appropriate, flow was switched to

plasma for 1.5 h to create a conditioning film. Thereafter, flow was switched for 30 min to

buffer for removal of all remnants of plasma from the tubes and the chamber, and then to the

bacterial suspension which was circulated through the system for 4 h.

Figure 3 is an example of the deposition kinetics of one strain, S. epidermidis 242. The

presence of a conditioning film on silicone rubber had a reducing effect on the adhesion of all

strains studied. The reduction of the initial deposition rate, j0, and the adhering numbers in a

stationary end-point varied between 77-97% and 55-98%, respectively, depending on the

strain used.


33

Time (103 sec)

0 2 4 6 8 10 12 140

2

4

6

8

10

12

14

Bac

teri

a (1

06 cm

-2)

Figure 3. Deposition kinetics of S. epidermidis 242 to silicone rubber with (triangles) and without (circles) a

plasma coating. For details, see Van der Mei et al. [27].

Adhesion and surface growth of Staphylococcus epidermidis on materials with various

wettabilities

Time (h)

0 5 10 15 20 25

Bac

teri

a (1

06 cm

-2)

0

20

40

60

80

100

Figure 4. The number of S. epidermidis HBH2 102 during deposition and surface growth on silicone rubber

(triangles) and glass (circles) [30].

Chapter 2

34

The adhesion and surface growth of Staphylococcus epidermidis HBH2 102 was determined

on different materials with varying water contact angles between 15° and 110° [30]. The flow

chamber temperature was kept at 36°C during the experiment. The initial adhesion rate was

determined in PBS for 1 h, the flow chamber was washed with PBS for 15 min and

subsequently was switched to 20 times diluted tryptone soya broth (TSB) in PBS. During 24 h

surface growth was followed. Due to its grape forming mode of growth, individual

staphylococci are not easily counted and the numbers of adhering bacteria were derived from

the measurement of the surface coverage of the biofilm. Figure 4 shows the numbers of

bacteria on silicone rubber and glass. The desorption of biofilm bacteria was determined and

expressed as the fraction of the biofilm that detaches per minute (kdes). The generation time

(g) can be calculated using a modification of the mathematical model for microbial biofilm

growth by Barton et. al. [31]:

( )ides

gti tknn ∆−= ∆ /

0 2 (1)

in which ni is the number of adhering bacteria after time i x ∆t and n0 is the bacterial number

at the start of the growth phase.

Figure 5 gives the generation time of this staphylococcal strain on the different

materials as a function of their water contact angle, revealing a relationship between

substratum wettability and surface growth.

Water contact angle (degrees)

0 20 40 60 80 100 120

Gen

erat

ion

time

(min

)

20

40

60

80

100

120

Figure 5. The generation time of S. epidermidis HBH2 102 during surface growth on different materials as a

function of their water contact angle [30].


35

Detachment of Pseudomonas aeruginosa from contact lenses by ophthalmic solutions

To study the efficacies of two contact lens cleansing solutions, first a biofilm was formed on a

contact lens quarter and mounted in the parallel plate flow chamber [16]. For this purpose, a

suspension of P. aeruginosa no. 3 in saline supplemented with 2 % TSB was circulated

through the system for 20 h to allow adhesion and surface growth. Then flow was switched

for 1 h to saline to remove non-adhering bacteria from the system and subsequently 8 ml of an

ophthalmic solution or saline (control), followed by 16 ml saline to clean the chamber of the

solution, was perfused through the flow chamber. To study renewed bacterial adhesion, flow

with saline was continued for 30 min, after which a new suspension of freshly cultured

bacteria was circulated through the system for 4 h. After removal of non-adhering bacteria

with saline for 1 h, another dose of ophthalmic solution was applied. Finally a liquid-air

interface was led over the surface. Figure 6 illustrates the number of adhering P. aeruginosa

on a contact lens with a tear film during one complete experiment. The ophthalmic solution

clearly decreases the number of bacteria while the liquid-air interface does not yield a

significant effect.

Time (min)

0 100 200 1200 1300 1400 1500 1600 1700 1800

Bac

teri

a (1

06 cm

-2)

0

1

2

3

4

5

6

7

R II

R I

air-bubble

Figure 6. Number of adhering P. aeruginosa no. 3 to a contact lens with an adsorbed tear film. After the

adhesion and surface growth, the surface was rinsed by a detergent mixture (R I), followed by a second adhesion

phase and a second rinse (R II). Finally an air bubble was passed over the surface. For details see Landa et.al.

[16].

Chapter 2

36

Advantages and disadvantages of the system

The major advantages of the system outlined are controlled shear and mass transport; a high

data density in time; and the avoidance of air-liquid interface passages over the adhering

microorganisms. Furthermore, the in situ observation offers the great advantage that all events

in initial biofilm formation, including adhesion and growth can be followed in time and that

the fate of an individual microorganism in the biofilm can be studied.

A disadvantage of the system is that biofilm formation is only viewed in two

dimensions, and consequently only initial biofilm formation can be studied. As the biofilm

thickness extends to above one layer, the events are not clearly visible anymore. To study

more mature biofilms, a three-dimensional viewing system is needed, as provided by scanning

confocal laser microscopy combined with three-dimensional image analysis software [34].

Unfortunately this method is invasive due to the fluorescent staining, and biofilm processes

cannot be followed in situ.

References

1. Walker, J.T., Hanson, K., Caldwell, D. and Keevil, C.W. (1998). Scanning confocal laser microscopy

study of biofilm induced corrosion on copper plumbing tubes. Biofouling 12, 333-44.

2. Notermans, S., Dormans, J.A. and Mead, G.C. (1991). Contribution of surface attachment to the

establishment of microorganisms in food processing plants: a review. Biofouling 5, 21-36.

3. Flint, S.H., Bremer, P.J. and Brooks, J.D. (1997). Biofilms in dairy manufacturing plant description.

Biofouling 11, 81-97.

4. Marsh, P. D. and Martin, M. V. (1992). Oral Microbiology, Chapman and Hall, London.




Science 237, 1588-95.

7. Wolfaardt, G.M., Lawrence, J.R., Robarts, R.D., Caldwell, S.J. and Caldwell, D.E. (1994).

Multicellular organization in a degradative biofilm community. Applied and Environmental Microbiology

60, 434-46.

8. Reid, G., Bruce, A.W., McGroarty, J.A., Cheng, K.J. and Costerton, J.W. (1990). Is there a role for

lactobacilli in prevention of urogenital and intestinal infections? Clinical Microbiology Reviews 3, 335-44.

9. Escher, A. and Characklis, W.G. (1990). Modeling the initial events in biofilm accumulation. In Biofilms .

Edited by Characklis, W.G. and Marshall, K.C. (Wiley, New York) p. 445-86.


37

10. Vieira, M.J., Melo, L.F. and Pinheiro, M.M. (1993). Biofilm formation: hydrodynamic effects on internal

diffusion and structure. Biofouling 7, 67-80.





12. Van Oss, C. J. (1994). Polar or Lewis acid-base interactions. In: Interfacial forces in aqeous media. Edited

by Van Oss, C.J. (Marcel Dekker, New York) p. 18-46.

13. Golomb, G. and Spigelman, A. (1991). Prevention of bacterial colonization on polyurethane in vitro by

incorporated antibacterial agent. Journal of Biomedical Materials Research 25, 937-52.



15. Rittmann, B.E. (1989). Detachment from biofilms. In Structure and Function of Biofilms. Edited by

Characklis, W.G. and Wilderer, P.A. (Wiley, New York) p. 49-58.

16. Landa, A.S., Van der Mei, H.C., Van Rij G. and Busscher, H.J. (1998). Efficacy of ophthalmic

solutions to detach adhering Pseudomonas aeruginosa from contact lenses. Cornea 17, 293-300.

17. Anwar, H. and Costerton, J.W. (1992). Effective use of antibiotics in the treatment of biofilm-associated

infections. ASM news 58, 665-8.

18. Busscher, H.J., Bos, R. and Van der Mei, H.C. (1995). Initial microbial adhesion is a determinant for the

strength of biofilm adhesion. FEMS Microbiology Letters 128, 229-34.

19. Busscher, H.J. and Van der Mei, H.C. (1995). Use of flow chamber devices and image analysis methods

to study microbial adhesion. Methods in Enzymology 253, 455-76.

20. Van Hoogmoed, C.G., Van der Mei, H.C. and Busscher, H.J. (1996). The influence of calcium on the

initial adhesion of S. thermophilus to stainless steel under flow studied by metallurgical microscopy.

Biofouling 11, 167-76.

21. Sjollema, J., Busscher, H.J. and Weerkamp, A.H. (1989). Experimental approaches for studying

adhesion of microorganisms to solid substrata: applications and mass transport. Journal of Microbiological

Methods 9, 79-90.

22. Wit, P.J. and Busscher, H.J. (1998). Application of an artificial neural network in the enumeration of

yeasts and bacteria adhering to solid substrata. Journal of Microbiological Methods 23, 281-290.

23. Kooten, T.G., Schakenraad, J.M., Van der Mei, H.C. and Busscher, H.J. (1993). Influence of pulsatile

flow on the behaviour of human fibroblasts adhered to glass. Journal of Biomaterial Science: Polymer

Edition 4, 601-14.

24. Pratt-Terpstra, I.H., Weerkamp, A.H. and Busscher, H.J. (1987). Adhesion of oral streptococci from a

flowing suspension to uncoated and albumin-coated surfaces. Journal of General Microbiology 133, 3199-

206.

25. Landa, A.S., Van der Mei, H.C. and Busscher, H.J. (1996). A comparison of the detachment of an

adhering oral streptococcal strain stimulated by mouthrinses and a pre-brushing rinse. Biofouling 9, 327-39.

Chapter 2

38

26. Velraeds, M.M.C., Van der Mei, H.C., Reid, G. and Busscher, H.J. (1997). Inhibition of initial adhesion

of uropathogenic Enterococcus faecalis to solid substrata by an adsorbed biosurfactant layer from

Lactobacillus acidophilus. Urology 49, 790-4.

27. Van der Mei, H.C., Van de Belt-Gritter, B., Reid, G., Bialkowska-Hobrzanska, H. and Busscher, H.J.

(1997). Adhesion of coagulase-negative staphylococci grouped according to physico-chemical surface

properties. Microbiology 143, 3861-70.

28. Zdanowski, Z., Ribbe, E. and Schalen, C. (1993). Influence of some plasma proteins on in vitro bacterial

adherence to PTFE and Dacron vascular prostheses. APMIS 101, 926-32.

29. Cheung, A.L., Krishnan, M., Jaffe, E.A. and Fischetti, V.A. (1991). Fibrinogen acts as a bridging

molecule in the adherence of Staphylococcus aureus to cultured human endothelial cells. The Journal of

Clinical Investigation 87, 2236-45.

30. Gottenbos, B., Van der Mei, H.C. and Busscher, H.J. (2000). Initial adhesion and surface growth of

Staphylococcus epidermidis and Pseudomonas aeruginosa on biomedical polymers. Journal of Biomedical

Materials Research 50, 208-14.



32. Habash, M.H., Van der Mei, H.C., Reid, G. and Busscher, H.J. (1997). Adhesion of Pseudomonas

aeruginosa to silicone rubber in a parallel plate flow chamber in the absence and presence of nutrient broth.

Microbiology 143, 2569-74.

33. Busscher, H.J., Doornbusch, G.I. and Van der Mei, H.C. (1992). Adhesion of mutans streptococci to

glass with and without a salivary coating as studied in a parallel-plate flow chamber. Journal of Dental

Research 71, 491-500.

34. Caldwell, D.E., Korber, D.R. and Lawrence, J.R. (1993). Analysis of biofilm formation using 2D vs 3D

digitial imaging. Journal of Applied Bacteriology, Symposium Supplement 74, 52S-66S.

39

Initial adhesion and surface growth of Staphylococcus

epidermidis and Pseudomonas aeruginosa on

biomedical polymersBart Gottenbos, Henny C. van der Mei and Henk J. Busscher

The infection risk of biomaterial implants varies between different materials and is determined

by an interplay of adhesion and surface growth of the infecting organisms. In this study, we

compared initial adhesion and surface growth of Staphylococcus epidermidis HBH2 102 and

Pseudomonas aeruginosa AK1 on poly(dimethylsiloxane), Teflon, polyethylene, polypropylene,

polyurethane, poly(ethylene terephthalate), poly(methyl methacrylate) and glass. Initial adhesion

was measured in situ in a parallel plate flow chamber with microorganisms suspended in

phosphate buffered saline, while subsequent surface growth was followed in full and in 20 times

diluted growth medium. Initial adhesion of both bacterial strains was similar to all biomaterials.

In full growth medium, generation times of surface-growing S. epidermidis ranged from 17 to

38 min with no relation to wettability, while in diluted growth medium generation times

increased from 44 to 98 min with increasing surface wettability. For P. aeruginosa no

influence of surface wettability on generation times was observed, but generation times

increased with decreasing desorption rates, maximal generation times being 47 min and

minimal values down to 30 min. Generally generation times of adhering bacteria were shorter

than of planktonic bacteria. In conclusion, surface growth of initially adhering bacteria is

influenced by biomaterial surface properties to a greater extent than initial adhesion.

Reproduced with permission of John Wiley & Sons, Inc. from

Journal of Biomedical Materials Research 50 (2000) 208-14

3

Chapter 3

40

Introduction

Biomaterial-centered infections are serious complications associated with the use of biomaterial

implants and devices. After adhesion and growth of infectious microorganisms on a biomaterial

surface, the biofilm mode of growth protects the organisms against the host defense system and

antibiotics [1]. Consequently, an infected biomaterial implant has to be replaced, at the expense

of considerable costs and patients discomfort [2]. Clinically, there appear to be great differences

in infection risk between different biomaterials. An expanded polytetrafluoroethylene (e-PTFE)

abdominal wall prosthesis, for example, is highly susceptible to infection, while the same

prosthesis made of polypropylene mesh demonstrates an impressive resistance against infections

[3].

The formation of an infectious biofilm is initiated by transport and adhesion of micro-

organisms to the surface of the implant or device [4]. Initial microbial adhesion is extensively

studied and generally believed to depend on the physico-chemical properties of the microbial and

biomaterial surfaces [5]. After initial adhesion, surface growth of adhering microorganisms leads

to the formation of a biofilm. Although this process is evidently significant in the pathogenesis of

biomaterial-centered infections, there are only few studies on surface growth of adhering

microorganisms, i.e. growth of those organisms in direct contact with the biomaterial surface [6].

Yet, these microorganisms play a pivotal role in biofilm formation as they link the entire biofilm

to the biomaterial surface [7].

Recently, Barton et al. [8] compared the initial surface growth rate of Staphylococcus

epidermidis, Pseudomonas aeruginosa and Escherichia coli on different orthopedic implant

materials in a parallel plate flow chamber in whole growth medium. A correlation was found

between the generation time of P. aeruginosa and the free energy of adhesion of the organisms

for the different biomaterials. This correlation was not found for S. epidermidis and E. coli.

Habash et al. [9] studied adhesion of P. aeruginosa AK1 to silicone rubber in a parallel plate

flow chamber from buffer and buffer, supplemented with 2 % nutrient broth. In broth

supplemented buffer, a steady increase of the number of adhering organisms was found also after

several hours, whereas in buffer stationary numbers of adhering bacteria were found after 1 h.

The steady increase was attributed to surface growth and corresponded with a higher initial

deposition rate in broth supplemented buffer as compared to in buffer only.

The aim of this study was to compare the initial adhesion and surface growth of two

clinically relevant strains of S. epidermidis and P. aeruginosa on different biomaterials, as listed

Initial adhesion and surface growth on biomedical polymers

41

Table 1. Proprietary names and usage of some common biomedical polymers [10].

Polymers Abbreviation Proprietary Names Implants or Devices

Poly(dimethyl

siloxane)

SR Silastic Pacemakers

Arteriovenous shunts

Intravascular devices

Cerebrospinal fluid shunts

Urological catheters

Peritoneal dialysis catheters

Mammary prostheses

Voice prostheses

Polyethylene PE Polythene, Alkathene,

Carlona, Marlex, Alathon


Cerebrospinal fluid shunts

Orthopedic implants

Poly(ethylene

terephthalate)

PET Dacron, Terlene, Mirafil,

Mersilene, Mylar, Melinex

Prosthetic heart valves

Left ventricular assist device

Total artificial heart

Vascular grafts



Peritoneal dialysis catheters

Mammary prostheses

Poly(methyl

methacrylate)

PMMA Acrylic, Diakon, Lucite,

Oroglas, Perspex, Plexiglass

Bone cement

Cranioplastic implants

Intraocular artificial lens

Polypropylene PP Propathene, Prolene,

Surgilene


Abdominal wall prosthesis

Poly(tetrafluoro

ethylene)*

PTFE Teflon, Fluon Vascular grafts



Abdominal wall prosthesis

Polyurethane PUR Estane, Lycra, Etheron,

Polyfoam, Biomer, Tecoflex,

Pellethane, Texin,

Avcothane, Electrolour

Left ventricular assist device

Total artificial heart

Pacemakers


Mammary prostheses

* Because of the turbidity of PTFE, poly(tetrafluoroethylene-co-hexafluoropropylene) (FEP) was used in the

present experiments.

Chapter 3

42

in Table 1 together with their possible application as an implant. Furthermore, glass was included

as it constitutes an extremely hydrophilic surface.

Materials and methods

Strains and growth conditions

S. epidermidis HBH2 102, isolated from the skin, and P. aeruginosa AK1, a uropathogenic

isolate, were used in this study. First, a strain was streaked and grown overnight at 37°C from a

frozen stock on a blood agar plate. The plate was then kept at 4°C, never longer than a week.

Several colonies were used to inoculate 5 ml of tryptone soya broth (TSB, OXOID, Basingstoke,

England) for S. epidermidis or nutrient broth (NB, OXOID) for P. aeruginosa in phosphate

buffered saline (PBS) that was incubated at 37°C in ambient air for 24 h. From this preculture 1

ml was used to inoculate a second culture (150 ml TSB in PBS for S. epidermidis or 100 ml NB

in PBS for P. aeruginosa) that was grown for 17 h. The bacteria from the second culture were

harvested by centrifugation (S. epidermidis at 5000 and P. aeruginosa at 10,000 g) for 5 min at

10°C and washed twice with sterile Millipore-Q water. Subsequently, the bacteria were sonicated

on ice (4 times 10 s with S. epidermidis and 3 times 10 s with P. aeruginosa) in sterile PBS. The

suspension was diluted in sterile PBS to a concentration of 3 x 108 cells ml-1.

In order to determine the generation time of the bacteria in suspension, bacteria of three

independently grown cultures were suspended in full or 20 times diluted growth medium in PBS

and growth curves were made at 36°C by optical density measurements at 600 nm.

Biomaterials

Sheets of implant grade poly(dimethylsiloxane) (Medin, Groningen, The Netherlands) were

heated for 4 h at 180°C to remove volatile impurities. Polyethylene, polypropylene and

poly(ethylene terephthalate) were obtained from Goodfellow (Cambridge, UK). Polyurethane

(pellethane 2363-75D) was kindly provided by Dow Benelux (Terneuzen, The Netherlands).

Further poly(tetrafluoroethylene-co-hexafluoropropylene) (Fluorplast, Raamsdonkveer, The

Netherlands), poly(methyl methacrylate) (Vink Kunststoffen, Didam, The Netherlands) and glass

were used. The samples were cleaned in a 2% RBS 25 (Omniclean, Breda, The Netherlands)

detergent solution under simultaneous sonication and thoroughly rinsed in demineralized water,

sterilized in 70% ethanol and finally washed with sterile Millipore-Q water.


43

The wettability of the materials was determined by water contact angle measurement at

room temperature with an image analyzing system, using the sessile drop technique. Each value

was obtained by averaging five droplets on one sample.

The parallel plate flow chamber, image analysis, adhesion and surface growth assay

The flow chamber (dimensions; l x w x h = 76 x 38 x 0.6 mm) and image analysis system have

been described in detail [11]. Images were taken from the bottom plate (58 x 38 mm) of the

parallel plate flow chamber which consisted of a thin square (15 x 15 mm) of the biomaterial to

be used affixed centrally into the groove (15 x 15 mm) of a thicker (2.0 mm) perspex plate. The

depth of the groove was adapted to the thickness of the biomaterial in such way that the materials

surface was on the same height as the surface of the perspex plate. In the studies on glass and

PMMA, the entire bottom plate was made of glass or PMMA, respectively. The top plate of the

chamber was always made of glass. The chamber was heat sterilized as a whole, except for the

perspex plate and biomaterials, which were sterilized by 70 % ethanol. The flow chamber was

equipped with heating elements and kept at 36°C throughout an experiment. Deposition was

observed with a CCD-MXRi camera (High Technology, Eindhoven, The Netherlands) mounted

on a phase contrast microscope (Olympus BH-2) equipped with a 40 x ultra long working

distance (Olympus ULWD-CD Plan 40 PL) and a 4 x objective. The camera was coupled to an

image analyzer (TEA, Difa, Breda, The Netherlands).

Prior to each experiment, all tubes and the flow chamber were filled with sterile PBS,

taking care to remove all air bubbles from the system. Once the system was filled, and prior to

the addition of the bacterial suspension, the fluid was allowed to flow through the system at a

flow rate of 0.025 ml s-1 corresponding with a shear rate of 10 s-1 for 60 min, while the flow

chamber was heated to 36°C, and subsequently switched to a bacterial suspension of 3 x 108 cells

ml-1 in PBS at the same flow rate. The bacterial suspension was perfused through the system for

1 h without recirculation and images were obtained continuously and analyzed real-time. The

initial increase in the number of adhering bacteria over time was expressed in so-called initial

deposition rate, i.e. the increase in the number of adhering bacteria per unit area and time.

Following 1 h perfusion of the flow chamber with bacterial suspension, flow was

switched to buffer without bacteria to remove unbound organisms from the tubes and the flow

chamber under the same flow rate for 15 min. Subsequently, flow was switched to either full or

20 times diluted growth medium in PBS. The number of bacteria detaching upon sudden

exposure to growth medium was determined and expressed as a percentage of the bacteria

Chapter 3

44

adhering prior to introduction of the medium. Growth medium was perfused through the system

at the same flow rate for 24 h without recirculation and images were recorded every 20 min for

S. epidermidis using a 4x objective and for P. aeruginosa every 3 min with a 40x objective.

Biofilm growth was mathematically analyzed by [8]

( )ides

gti tknn ∆−= ∆ /

0 2 (1)

in which ni is the number of adhering bacteria after time i x ∆t, n0 is the bacterial number at

the start of the growth phase, g is the generation time and kdes is the desorption rate constant.

For S. epidermidis, the mathematical analysis was based on surface coverage by the biofilm. The

area of desorbed biofilm during each time interval was divided by the surface coverage of the

total biofilm at the beginning of the time increment yielding the desorption rate constant (kdes).

Subsequently the generation time was regressed from equation 1 by performing a nonlinear

least-squares fit of the logarithmic growth phase of the growth curve. For P. aeruginosa, the

mathematical analysis was based on numbers of adhering bacteria and the generation time could

be directly obtained from the recorded images as cell division was clearly visible. Generation

time was determined for at least 30 individual bacteria and averaged. From equation 1, the

desorption rate constant kdes was regressed by fitting the logarithmic part of the growth curve.

Furthermore, for both bacteria a so-called “biofilm doubling time” (tdouble) was determined, i.e.

the time after which the number of bacteria on the surface had doubled. The growth

characteristics were determined after the lag time. After 24 h of surface growth, the surface

coverage of the biofilm was determined for 15 different fields of view for each bacterial strain.

Experiments in full medium were done in single fold, while those in diluted medium were

done in duplicate.

Results

Wettability of the biomaterials

In Table 2, the water contact angles (θw) on the materials employed are shown to cover a

broad range of wettabilities, varying in water contact angle from 20 to 111 degrees.


45

Table 2. Water contact angles, initial adhesion rate and numbers adhering after 1 h during deposition of bacteria

from PBS to biomedical polymers

S. epidermidis HBH2 102 P. aeruginosa AK1

Biomaterial θw (degrees) j0

(102 cm-2 s -1)

n1h

(106 cm-2)

j0

(102 cm-2 s -1)

n1h

(106 cm-2)

SR 111 18 6.4 2.9 0.7

FEP 108 21 5.8 1.1 0.3

PE 98 19 6.7 1.9 0.9

PP 94 18 6.0 3.6 0.9

PUR 78 17 5.7 3.8 0.7

PMMA 75 21 6.4 3.1 0.6

PET 74 22 7.4 3.2 0.7

Glass 20 19 5.4 2.6 0.4

Initial bacterial adhesion

During the first part of the experiments, the initial adhesion of S. epidermidis HBH2 102 and

P. aeruginosa AK1 onto the different materials was determined in PBS, i.e. in the absence of

growth. The initial deposition rates (j0) and total numbers of adhered bacteria after 1 h (n1h)

are shown in Table 2 as well. Initial deposition of the S. epidermidis strain is 5 to 20 times

faster than of P. aeruginosa AK1, with little variation between the different materials. Most

prominent is the low initial deposition rate of P. aeruginosa on FEP, as compared with the

other materials. Similarly, also the number of bacteria adhering after 1 h are much more strain

dependent rather than substratum dependent.

Surface growth of adhering bacteria

Examples of the surface growth of S. epidermidis in 20 times diluted TSB on PE, PP and PET

are shown in Figure 1. At the onset of the experiment (time 0), flow was switched from buffer

to 20 times diluted TSB and after approximately 7 h, the surface coverage of the biofilm

started to increase exponentially, while the bacteria grew in aggregates expanding over the

biomaterial surface. The lag time was 4 h in full TSB for S. epidermidis. For P. aeruginosa

the lag time was 1 h for both media.

Chapter 3

46

Time (h)

0 5 10 15 20

Surf

ace

cove

rage

(%

)

0.1

1

10

100

Figure 1. Surface coverage by S. epidermidis HBH2 102 biofilm during flow of 20 times diluted TSB through

the parallel plate flow chamber on PP (circles), PET (squares) and PE (triangles).

Table 3 reports the results of the surface growth experiments of S. epidermidis HBH2

102. Desorption upon first exposure to the medium was less than 10%, except for FEP in full

TSB, from which a large percentage of adhering bacteria desorbed. In full growth medium the

desorption rate constants during growth (kdes) were highest on PE, PP, PUR and PET, while in

diluted medium these were highest on glass, SR, FEP, and PMMA. Biofilm doubling times as

well as generation times were much shorter in full than in 20 times diluted TSB. In full

medium these were shortest on SR, PE and PUR and longest on FEP, PMMA and PET. In

diluted growth medium biofilm doubling times ranged between 68 and 258 min with the

longest doubling times occurring on PP and glass. Here generation times of surface-growing

bacteria exceeded the generation time of planktonically growing bacteria on PP, PMMA, PET

and glass. In full TSB surface growth appeared faster than growth of planktonic bacteria, for

which a generation time of 52 min was found. After 24 h, most surfaces were completely

covered with a biofilm in full growth medium, while in diluted medium surface coverage after

24 h of surface growth ranged between 6 and 80%.


47

80

ND

**

46 6 35 27 21 7

Cov

erag

e 24

h (%

)

100

70 100

100

100

97 100

100

67 44 47 55 97 53 69 78 98

g (m

in)

52 20 36 17 30 19 38 34 27

68 75 72 193

82 118

134

258

t dou

ble (

min

)

21 42 20 36 22 43 43 30

1.3

1.4

0.7 1 1 1.1 1 1.6

k des

(10-2

min

-1)

0.6

0.6

1.1 1 1.3

0.5 1 0.6

0 0 4 3 0 2 3 8

Des

orpt

ion

Med

ium

(%)

6 40 0 4 0 7 0

ND

**

Bio

mat

eria

l

Plan

kton

ic

SR FEP

PE PP PUR

PMM

A

PET

Gla

ss

Tab

le 3

. Gro

wth

cha

ract

eris

tics

of S

. epi

derm

idis

HB

H2

102

on b

iom

edic

al p

olym

ers

in fu

ll (l

eft c

olum

ns) a

nd 2

0 tim

es d

ilute

d T

SB (r

ight

col

umns

)*.

* E

xper

imen

ts in

20

times

dilu

ted

TSB

wer

e ca

rrie

d ou

t in

dupl

icat

e an

d co

inci

ded

with

in 2

0% fo

r kde

s, 26

% fo

r tdo

uble, 2

0% fo

r g a

nd 4

0% fo

r cov

erag

e af

ter 2

4 h.

** N

D =

not

det

erm

ined

.

Chapter 3

48

64 54 68 58 60 61 50 54

Cov

erag

e 24

h (%

)

ND

**

61

ND

**

67 68

ND

**

51 65

57 43 39 46 44 44 43 47 47

g (m

in)

40 29 31 37 43 41 30 43 30

58 81 59 55 54 65 52 75

tdou

ble

(min

)

51 44 45 45 43 46 44 58

1.2

2.7

0.7

0.7

0.6

1.3

0.3

1.4

k des

(10-2

min

-1)

2.7

1.5

0.8

0.2

0.2

1.8

0.1

3.2

40 38 5 3 2 29 3 30

Des

orpt

ion

Med

ium

(%)

40 29 0 5 8 70 3 39

Bio

mat

eria

l

Plan

kton

ic

SR FEP

PE PP PUR

PMM

A

PET

Gla

ss

Tab

le 4

. Gro

wth

cha

ract

eris

tics

of P

. aer

ugin

osa

AK

1 on

bio

med

ical

pol

ymer

s in

full

(lef

t col

umns

) and

20

times

dilu

ted

NB

(rig

ht c

olum

ns)*

* E

xper

imen

ts in

20

times

dilu

ted

NB

wer

e ca

rrie

d ou

t in

dupl

icat

e an

d co

inci

ded

with

in 3

9% fo

r kde

s, 9%

for t

doub

le, 7

% fo

r g a

nd 1

5% fo

r cov

erag

e af

ter 2

4 h.

** N

D =

not

det

erm

ined

.


49

Surface growth of P. aeruginosa AK1 proceeded along a different pattern than that of

S. epidermidis, as adhering P. aeruginosa elongate, divide and split up during growth, rather

than forming aggregates. Table 4 presents the results of the mathematical analysis of P.

aeruginosa biofilm formation. Desorption upon exposure to medium is low (0-8%) on PE, PP,

PUR and PET, while on the other materials desorption is high (29-70%). Desorption rate

constants during growth on the different biomaterials vary accordingly. The biofilm doubling

time in full NB is highest on glass and SR. In full medium, biofilm doubling times varied less

amongst the different biomaterials and were shorter than in diluted medium. Generally,

bacteria grew faster on the biomaterial surfaces than planktonically. The surface coverage

after 24 h varied between 50 and 68%.

Discussion

In this study we evaluated the extended use of the parallel plate flow chamber to study initial

adhesion and surface growth of bacteria onto various biomaterial surfaces. It was shown that the

system is suitable to measure bacterial growth and desorption rates during surface growth on

various materials. A combined adhesion and growth experiment with this system is novel.

Combination of adhesion experiments with subsequent growth measurement is clinically much

more relevant, as implant infections usually result from surface growth of only a few adhering

microorganisms.

Bacterial adhesion is often found to differ between materials with different chemical

composition or hydrophobicity [12]. In our study little variation was seen in initial adhesion to

different materials varying in chemical composition and hydrophobicity. This can probably be

explained by the high ionic strength of PBS, in which the experiments were performed, as

differences in electrostatic interactions become negligible in high ionic strength solutions [13].

The results obtained during surface growth of S. epidermidis HBH2 102 for FEP and PP

correspond strikingly well with clinical experiences obtained with abdominal wall patches made

out of these biomaterials. Growth rate on PP under low nutrient conditions is 2 times lower than

on FEP and on the other more hydrophobic biomaterials. This corresponds to the low incidence

of infections of PP abdominal wall prostheses in comparison to e-PTFE prostheses [3]. In full

TSB, these differences were not seen, but this high nutrient level likely does not correspond to

the clinical situation. As most biomaterial-centered infections are caused by staphylococci [2],

these data suggest that growth rate is an important factor for the pathogenesis of staphylococcal

Chapter 3

50

infections, possibly because slow growth of the infecting organisms changes the “equation of

infection” in favor of the host [14].

Several studies have been done on the relation between materials surface properties and

growth of adhering bacteria. Van Loosdrecht et al. [15] concluded that adhesion of bacteria does

not directly influence their metabolism and growth yield and changes in growth rate of adhering

bacteria with respect to planktonically growing bacteria are suggested to be a result of nutrient

adsorption. Depending on the amount of adsorbed nutrients and whether adsorption is easily

reversed, growth rates of adhering bacteria can decrease or increase with respect to their

planktonic counterparts.

Interestingly, generation times of S. epidermidis in 20 times diluted TSB decrease

linearly on the different materials with increasing contact angles (excluding PP, see Figure 2).

Water contact angle (degrees)

0 20 40 60 80 100 120

Gen

erat

ion

time

(min

)

0

20

40

60

80

100

120

Figure 2. Relation between generation times during surface growth of S. epidermidis HBH2 102 and the water

contact angles on the biomaterial surfaces. The generation times in full TSB are shown as circles, while the

squares give the generation time in 20 times diluted TSB. An arrow indicates the PP data point in diluted TSB.

In full TSB the generation times were not systematically influenced by the surface

hydrophobicity. Likely, the availability of adsorbed nutrients is different on hydrophilic than on

hydrophobic biomaterials, while the abundant nutrient availability in full growth medium is

sufficient to ensure equal growth opportunities on all biomaterial surfaces.


51

With P. aeruginosa no relations were found between surface growth and biomaterials

wettability. Differences between the generation times on the different biomaterials were largest

in full NB and only minor in 20 times diluted NB. On those biomaterials, showing slow growth

of adhering bacteria, desorption of bacteria upon exposure to growth medium was low (see Table

4), while furthermore also bacterial generation times decreased with increasing desorption rate

constants in full NB (see Figure 3). In diluted NB this relation between generation time and

desorption rate constant was less evident. Tentatively, and interpreting low desorption as

indicative for a strong adhesion force between bacteria and biomaterial surface, it is suggested

that stronger attraction between adhering bacteria and a surface hampers elongation and

division of the adhering bacteria and therewith surface growth. This suggestion is in line with

hypotheses of Harkes et al. [16] demonstrating that surface growth of adhering E. coli was slow

or even absent on positively charged surfaces, due to tight binding as compared with negatively

charged surfaces.

Desorption rate constant (min -1)

0.00 0.01 0.02 0.03

Gen

erat

ion

time

(min

)

28

30

32

34

36

38

40

42

44

46

48

Figure 3. Relation between generation times and desorption rate constants during surface growth of P.

aeruginosa AK1 in full NB (circles) and 20 times diluted NB (squares).

In conclusion, surface growth of bacteria adhering on biomaterial surfaces is affected by

physico-chemical interactions between nutrients, bacteria and the biomaterial surface. Surface

growth rates of S. epidermidis HBH2 102 and P. aeruginosa AK1 are different on different

Chapter 3

52

biomaterials and probably account for differences in clinical risk of biomaterial-centered

infections, whereas initial bacterial adhesion is relatively similar.

References

1. Anwar, H. and Costerton, J.W. (1992). Effective use of antibiotics in the treatment of biofilm-associated

infections. ASM News 58, 665-8.


Science 237, 1588-95.

3. Ten Cate-Hoedemaker, H.O. (1997). Prosthetic materials in abdominal surgery. In Biomedical materials

research in Groningen. Edited by Poelstra, K.A., Busscher, H.J. and Van Horn, J.R. (Uitgeverij Van Brug,

Groningen) p. 109-10.

4. Escher, A. and Characklis, W.G. (1990). Modeling the initial events in biofilm accumulation. In Biofilms .

Edited by Characklis, W.G. and Marshall, K.C. (Wiley, New York) p. 445-86.

5. Bos, R., Van der Mei, H.C. and Busscher, H.J. (1999). A physico-chemical approach towards initial

microbial adhesive interactions. FEMS Microbiology Reviews 23, 179-230.

6. Geesey, G.G. and White, D.C. (1990). Determination of bacterial growth and activity at solid-liquid

interfaces. Annual Review of Microbiology 44, 579-602.












12. An, Y.H. and Friedman, R.J. (1998). Concise review of mechanisms of bacterial adhesion to biomaterial

surfaces. Journal of Biomedical Materials Research 43, 338-48.

13. Bouttier, S., Han, K.G., Ntsama, C., Bellon-Fontaine, M.N. and Fourniat, J. (1994). Role of

electrostatic interactions in the adhesion of Pseudomonas fragi and Brochothrix thermosphacta to meat.

Colloids and Surfaces B: 2, 57-65.






53

15. Van Loosdrecht, M.M., Lyklema, J., Norde, W. and Zehnder, A.B. (1990). Influence of interfaces on

microbial activity. Microbiological Reviews 54, 75-87.



55

Antimicrobial effects of positively charged surfaces

on adhering Gram-positive and Gram-negative

bacteriaBart Gottenbos, Dirk W. Grijpma, Henny. C. van der Mei, Jan Feijen and Henk J. Busscher

The occurrence of biomaterials-centered infection is determined by an interplay of adhesion and

surface growth of the infecting organisms. In this paper, the antimicrobial effects on adhering

bacteria of a positively charged poly(methacrylate) surface (zeta potential +12 mV) are

compared with those of negatively charged poly(methyl methacrylate) (-12 mV) and a strong

negatively charged poly(methacrylate) (-18 mV) surface. Initial adhesion of Staphylococcus

aureus ATCC 12600, Staphylococcus epidermidis HBH2 102, Escherichia coli O2K2 and

Pseudomonas aeruginosa AK1 to these biomaterial surfaces was measured in a parallel plate

flow chamber in phosphate buffered saline, while subsequently adhering bacteria were allowed

to grow by perfusing the flow chamber with growth medium. On the positively charged surface

adhesion was fastest, but subsequent surface growth, however, was absent for the Gram-negative

strains. On the negatively charged surfaces, despite a slower initial adhesion, surface growth of

the adhering bacteria was exponential for both Gram-positive and Gram-negative strains. These

results suggest that positively charged biomaterial surfaces exert an antimicrobial effect on

adhering Gram-negative bacteria, but not on Gram-positive ones.

Partly reproduced with permission of Kluwer Academic Publishers from:

Journal of Materials Science: Materials in Medicine 10 (1999) 853-55

and partly accepted for publication in the Journal of Antimicrobial Chemotherapy

4

Chapter 4

56

Introduction

Infection is still the most common cause of biomaterial implant failure in modern medicine

[1,2], and despite the fact that more and more different implants are designed, like for

example the total artificial heart, mammary prostheses, different orthopedic implants, and a

total artificial larynx prosthesis, there is at present no solution to this problem other than

removing the implant [1]. Adhesion and subsequent surface growth of bacteria on biomedical

implants and devices causes the formation of a biofilm in which the so-called “glycocalix”

embeds the infecting bacteria, offering protection against the host immune system and

antibiotic treatment. As most bacteria carry a net negative surface charge [3], adhesion of

bacteria is discouraged on negatively charged surfaces, while it is promoted on positively

charged surfaces [4-6]. However, adhesion is only one of the first steps in the formation of a

biofilm infection [7] and in order for a biofilm to fully develop, the adhering bacteria have to

grow [8]. Surface growth of the initially adhering bacteria was found by Harkes et al. [9] to be

absent on positively charged poly(methacrylates) for Escherichia coli. Barton et al. [10]

found that surface growth of Pseudomonas aeruginosa correlated with the free energy of

adhesion, while no such correlation was found for Staphylococcus epidermidis and E. coli.

Recently, we reported [11] that growth of P. aeruginosa growing on biomaterial surfaces

decreased with an increasing strength of adhesion to the surface.

The aim of this study was to determine possible antimicrobial effects of a homologous

series of three methacrylate polymers and copolymers varying in surface charge on different

Gram-positive and Gram-negative bacterial strains. To this end, initial adhesion and

subsequent surface growth of Staphylococcus aureus, S. epidermidis, E. coli and P.

aeruginosa was measured in a parallel plate flow chamber.

Methods

Bacterial strains and growth conditions

S. aureus ATCC 12600, S. epidermidis HBH2 102, P. aeruginosa AK1 and E. coli O2K2 were

used in this study. First, a strain was streaked and grown overnight at 37°C from a frozen stock

on a blood agar plate. The plate was then kept at 4°C, never longer than a week. Several colonies

were used to inoculate 5 ml of tryptone soya broth (TSB, OXOID, Basingstoke, England) for the

staphylococci, nutrient broth (NB, OXOID) for P. aeruginosa or brain heart infusion (BHI,


57

OXOID) for E. coli in phosphate buffered saline (PBS). This preculture was incubated at 37°C in

ambient air for 24h and used to inoculate a second culture (150 ml TSB in PBS for the

staphylococci, 100 ml NB in PBS for P. aeruginosa or 100 ml BHI for E. coli) that was grown

for 18 h. The bacteria from the second culture were harvested by centrifugation (5 min, 10,000 g

for P. aeruginosa and 5000 g for the other strains) and washed twice with sterile Millipore-Q

water. Subsequently, the bacteria were suspended in sterile PBS, for S. epidermidis after

sonication on ice (3 x 10 s), to a concentration of 3 x 108 cells ml-1.

In order to determine the generation time of the bacteria in suspension, 1 ml of preculture

was suspended in 200 ml of the appropriate growth medium and growth curves were recorded at

37°C by optical density measurements at 600 nm. Generation times were derived from the

doubling times of the optical density values.

Polymer synthesis

For this study, homopolymers of methyl methacrylate (MMA, Merck) (PMMA) and copolymers

of MMA with either 15 mol% methacrylic acid (MAA, Aldrich) (PMMA/MAA 85/15) or 15

mol% trimethylaminoethyl methacrylate chloride (TMAEMA-Cl, 75%, Aldrich)

(PMMA/TMAEMA-Cl 85/15) were used (see structure formulas in Figure 1). Polymers were

synthesized as described previously [5,12] by radical polymerization of the monomers using

2,2’-azobis(methyl isobutyrate) as an initiator, synthesized from 2,2’-azobisbutyronitrile (Merck)

as described by Mortimer [13]. Solvents and MMA were distilled before use. The purity of the

polymers was checked by nuclear magnetic resonance.

O

ROMMA : R = CH3

MAA : R = H

TMAEMA-Cl : R = C3H6N+(CH3)3 Cl-

Figure 1. Structure formulas of the monomers used to synthesize the differently charged polymers used in this

study.

Polymer films were prepared on microscope glass slides (25x76x1 mm) and coverslips

(18x18x0.1 mm). Glass slides and cover slips were cleaned by immersion in a mixture of

hydrochloric acid (37%, p.a., Merck) and nitric acid (65%, p.a., Merck), ratio 3:1 (v/v) for 20 h.

After extensive rinsing with double deionized water and ethanol the slides and coverslips were

2 4

Chapter 4

58

dried in vacuo at 60 °C for 3 h. The glass slides and coverslips were silanized with n-

propyltrimethoxysilane (Aldrich) in case of PMMA and with γ-aminopropyltriethoxysilane

(Fluka) in case of PMMA/MAA. Silanization was performed by immersion of the glass surfaces

in a solution of the silane (1%, v/v) in toluene at room temperature for 2 h. Glass slides and

coverslips were rinsed with toluene and dried in vacuo at 60 °C for 3 h. The microscope glass

slides and coverslips were coated at one side by spin coating. Polymer solution (1 %, w/v) in

toluene for PMMA and in dimethylformamide for the copolymers was dispensed onto the slides

and coverslips covering the entire surface. Then, slides and slips were spun at 2000 rpm for 20 s

in case of PMMA and for 60 s in case of the copolymers. This procedure was repeated two times

to obtain film uniformity, assuring that all polymer films had a similar surface topography.

Finally, the polymer films were dried in vacuo at 60 °C for 18 h to remove all solvent remnants

and the coverslips were kept in a sterile environment for bacterial adhesion and growth

experiments, while the glass slides were used for surface characterization.

The chemical composition of the films was determined by X-ray photoelectron

spectroscopy (XPS) using a S-Probe spectometer (Surface Science Instruments, Mountain View

CA, USA). The elemental surface compositions were expressed in atomic %, setting %C + %O

+ %N + %Si + %Cl to 100%. Zeta potentials of the film surfaces were derived from the pressure

dependence of the streaming potentials employing rectangular platinum electrodes (5.0 x 25.0

mm) located at both ends of a parallel plate flow chamber [14], constituted by the glass slides, as

separated by an 0.2 mm Teflon gasket. Streaming potentials were measured over 7 h in PBS (pH

7.0) at ten different pressures ranging from 37.5 to 150 Torr and each pressure was applied for 10

s in both directions. Water contact angles were measured at room temperature with an image

analyzing system, using the sessile drop technique. Each value was obtained by averaging results

of at least three droplets on one sample.

Bacterial zeta potentials

Bacterial zeta potentials were derived from particulate microelectrophoresis [15,16]. Three

independently grown cultures of each strain were harvested and washed as described above.

Bacteria were re-suspended (5 x 107 cells ml-1) in sterile PBS (pH 7.0) and measurements were

taken immediately after re-suspending at 150 V using a Lazer Zee Meter 501 (PenKem, USA)

and converted into zeta potentials assuming the Helmholtz-Smoluchowski equation holds.


59

The parallel plate flow chamber, image analysis, adhesion and surface growth assay

The flow chamber (dimensions; l x w x h = 76 x 38 x 0.6 mm), image analysis system and

adhesion and surface growth assay have all been described in detail [17,18]. Images were taken

from the bottom plate (58 x 38 mm) of the parallel plate flow chamber, consisting of the spin-

coated microscope coverslip affixed centrally with double sided tape (0.06 mm thick) in a groove

(18 x 18 x 0.16 mm) made in a thicker (2.0 mm) PMMA plate. The top plate of the chamber was

made of glass. The chamber was heat sterilized as a whole, except for the PMMA plate, which

was sterilized in 70% ethanol. The coated coverslip was used without sterilization. The flow

chamber was equipped with heating elements and kept at 37°C throughout an experiment.

Deposition and surface growth was observed with a CCD-MXRi camera (High Technology,

Eindhoven, The Netherlands) mounted on a phase contrast microscope (Olympus BH-2)

equipped with a 40 x ultra long working distance (Olympus ULWD-CD Plan 40 PL). The

camera was coupled to an image analyzer (TEA, Difa, Breda, The Netherlands).

Prior to each experiment, all tubes and the flow chamber were filled with sterile PBS,

taking care to remove all air bubbles from the system. PBS was allowed to flow through the

system at a flow rate of 0.025 ml s-1 corresponding with a shear rate of 10 s-1 for 60 min, while

the flow chamber was heated to 37°C. Subsequently, flow was switched to a bacterial suspension

of 3 x 108 cells ml-1 in PBS at the same flow rate. The bacterial suspension was perfused through

the system for 1 h without recirculation. Following 1 h perfusion of the flow chamber with

bacterial suspension, flow was switched to buffer without bacteria to remove unbound organisms

from the tubes and the flow chamber under the same flow rate for 15 min. Finally, flow was

switched to growth medium, TSB for the staphylococci, NB for P. aeruginosa and BHI for E.

coli, and medium was perfused through the system at the same flow rate for 6 h without

recirculation. The experiments were performed in duplicate.

During the experiment images were recorded and automatically analyzed yielding the

number of adhering bacteria as a function of time. The initial increase in the number of adhering

bacteria over time was expressed in a so-called initial deposition rate, i.e. the increase in the

number of adhering bacteria per unit area and time. The division time of individually adhering

bacteria was monitored as well, yielding their generation time. The numbers of growing and

non-growing bacteria were determined starting with the image taken after 2 h of flow with

growth medium, and following the individual bacteria for the next 4 h. Subsequently, a

percentage number of growing bacteria relative to the number of adhering bacteria at 2 h was

Chapter 4

60

calculated. A desorption rate constant (kdes) was obtained by performing a non-linear least

squares fit of the increasing part of the growth curve using Barton et al. [10]:

( )i

desgt

gngi tknnn ∆−+= ∆ /0 2 (1)

in which ni is the number of bacteria after the ith time increment (∆t), nng is the determined

number of non-growing bacteria, ng0 is the number of growing bacteria at the beginning of the

logarithmic growth phase and g is the generation time. Desorption of non-growing bacteria

was negligible as compared to desorption of growing bacteria. When no growing bacteria

were present, kdes was determined by dividing the number of desorbed bacteria by the number

of adhering bacteria for each time increment after 2 h of flow with growth medium and

averaging these.

Table 1. Zeta potentials in PBS, water contact angles and chemical composition of various PMMA based

polymer films employed in this study.

PMMA/MAA PMMA PMMA/TMAEMA-Cl

Zeta potential (mV) -18 -12 +12

Water contact angle (°) 70 71 65

C (%) 69.3 70.6 70.9

O (%) 29.4 28.6 25.3

N (%) 0 0 1.9

Si (%) 1.3 0.9 0

Cl (%) 0 0 1.9

Results

Characterization of polymer films

Table 1 gives zeta potentials of the polymer films in PBS. Zeta potentials range from −18 mV

(PMMA/MMA) to +12 mV (PMMA/TMAEMA-Cl) and were stable over 7 h. Water contact

angles on the polymer films were typical of an intermediately hydrophobic surface and

showed no major variation with zeta potential, indicating that all results can be interpreted

without intervening influences of substratum hydrophobicity. XPS analyses indicate that the

positive charge originates from nitrogen containing groups, while the increased negative

charge is caused by oxygen containing groups on the modified acrylate surfaces.


61

Figure 2. Representative example images of surface growing S. aureus ATCC 12600 on negatively charged

PMMA/MAA (--), PMMA (-) and on positively charged PMMA/TMAEMA-Cl (+). The left series, taken after 2

h only represent adhesion, while the right series, taken after 4 h after introduction of the growth medium, include

surface growth. The bar represents 10 µm.

Bacterial zeta potentials

All bacterial strains studied here are negatively charged in PBS and their zeta potentials

amount −10 mV for S. aureus ATCC 12600, −8 mV for S. epidermidis HBH2 102, −16 mV

for E. coli O2K2 and −7 mV for P. aeruginosa AK1.

--2h

-2h

-4h

+4h

+2h

--4h

Chapter 4

62

Figure 3. Representative example images of surface growing P. aeruginosa AK1 on negatively charged

PMMA/MAA (--), PMMA (-) and on positively charged PMMA/TMAEMA-Cl (+). The left series, taken after 2

h only represent adhesion, while the right series, taken after 4 h after introduction of the growth medium, include

surface growth. The bar represents 10 µm.

Adhesion and surface growth

Figures 2 and 3 show example images during the growth phase of S. aureus and P. aeruginosa,

respectively. Note that more bacteria are adhering on the positively charged PMMA/TMAEMA-

Cl surface after 2 h, but after additional growth (4 h) this situation has changed, and most S.

aureus microcolonies are found on the PMMA surface. Proliferating P. aeruginosa are seen only

-2h

--4h

-4h

+2h

+4h

--2h


63

on the negatively charged surfaces and growth appears absent by comparison of the images taken

after 2 h and 4 h on the positively charged PMMA/TMAEMA-Cl surface. The numbers of

adhering bacteria during adhesion and surface growth on the charged methacrylates are shown

graphically in Figure 4. During the growth phase, proliferating staphylococci were present on all

surfaces from 1 h after the introduction of growth medium. On the negatively charged surfaces

most of the E. coli and P. aeruginosa cells were proliferating within 30 min. The numbers of E.

coli raised however slowly because most newly formed bacteria desorbed directly from these

surfaces.

P. aeruginosa AK1

Time (h)

0 1 2 3 4 5 6 7

S. aureus ATCC 12600

Bac

teria

(10

6 cm

-2)

0

4

8

12

16

20

-

+--

E. coli O2K2

Time (h)

0 1 2 3 4 5 6 7

Bac

teria

(10

6 cm

-2)

0

1

2

3

4

5

6

7

-+

--+

---

S. epidermidis HBH2 102

+

---

Figure 4. Example of the number of adhering Gram-positive (top) and Gram-negative (bottom) bacteria on

negatively charged PMMA/MAA (--) and PMMA (-) and on positively charged PMMA/TMAEMA-Cl (+) in a

parallel plate flow chamber. The dashed lines indicate the time period during which PBS was perfused through

the flow chamber prior to the introduction of growth medium.

Table 2 gives initial deposition rates (j0), percentages of growing bacteria after 2 h, generation

times (g) and desorption rate constants (kdes) of adhering bacteria. Initial deposition rates were

highest for the staphylococci and generally increased as the substrata became less negatively

charged. Under conditions of electrostatic attraction, as on PMMA/TMAEMA-Cl, initial

Chapter 4

64

adhesion rates were maximal. Initial adhesion rates of P. aeruginosa AK1 were lowest of all four

strains, but also increased as the electrostatic repulsion between the bacteria and the substratum

surface disappeared on PMMA/TMAEMA-Cl. Staphylococci grew on all substratum surfaces,

although the addition of negative and positive charge to PMMA decreased the relative number of

growing staphylococci by a factor of 4 and 2, respectively. Their generation times were

comparable with the ones of planktonic S. aureus ATCC 12600 (31 min) and S. epidermidis

HBH2 102 (46 min).

Table 2. Initial deposition rates (j0), percentages of growing bacteria, generation times (g) and desorption rate

constants (kdes) of Gram-positive and Gram-negative bacteria on PMMA/MMA (--), PMMA (-) and

PMMMA/TMAEMA-Cl (+) polymer films with different charge. The results are average values of duplicate

experiments, which coincided for the different parameters within 20, 40, 5 and 25%, respectively.

Strain Charge j0 (cm2 s -1) % growth g (min) kdes (10-5 s -1)

S. aureus ATCC 12600 -- 1600 8 41 8

- 1780 34 32 15

+ 2700 13 39 23

S. epidermidis HBH2 102 -- 1900 7 48 22

- 1360 26 50 12

+ 3630 15 48 17

E. coli O2K2 -- 240 91 24 70

- 720 59 22 42

+ 1720 0 no growth 2

P. aeruginosa AK1 -- 350 75 32 12

- 430 70 35 2

+ 660 0 no growth 1

Gram-negative rods only grew on negatively charged surfaces and their generation times also

compared with the generation times of planktonic bacteria, viz. 23 min for E. coli O2K2 and 43

min for P. aeruginosa AK1. However, Gram-negative rods did not grow on the positively

charged PMMA/TMAEMA-Cl. All strains showed desorption of adhering bacteria. For the

Gram-positive staphylococci, desorption rate constants on the positively charged material were

similar to those on the negatively charged materials, but the absence of growth for the Gram-

negative rods coincided with a very low desorption of E. coli and P. aeruginosa cells, whereas

the presence of negative charge increased the desorption rate.


65

Discussion

Current approaches to develop new biomaterials with a low risk of becoming infected once

implanted in the human body, are predominantly based on developing non-adhesive surfaces. It

is known already, that initial adhesion of coagulase-negative staphylococci [4,6] and E. coli [5] is

faster on positively charged PMMA/TMAEMA-Cl co-polymers than on negatively charged

PMMA and PMMA/MAA co-polymers, as found in this study. This is due to the absence of

repulsive electrostatic interactions between the negatively charged bacteria and the positively

charged PMMA/TMAEMA-Cl. The present results on the surface growth of the initially

adhering bacteria suggest, however, that adhesion and surface growth may be oppositely affected

by substratum charge. Positively charged surfaces may attract more bacteria, but this effect is

readily counterbalanced by the absence of any growth, at least for the Gram-negative strains

involved in this study. Positively charged surfaces may therefore be regarded as antimicrobial

surfaces for these organisms.

Previously, it has been demonstrated that whenever the binding strength of adhering P.

aeruginosa AK1 to substrata increases, this has a reducing influence on their surface growth

[11]. Complete inhibition of growth, as found here for Gram-negative rods, possibly indicates

that elongation of adhering bacteria, necessary for cell division, is impeded by strong binding

through the attractive electrostatic interactions. In addition, as soluble quaternary ammonium

salts have been known for a long time to exhibit antimicrobial activity [19] through interaction

with the bacterial cell membrane [20], the quaternary ammonium groups of the positively

charged polymer, although insoluble, may disrupt the cell membrane of the Gram-negatives.

Gram-positive bacterial strains have a comparatively thicker and more rigid peptidoglycan layer

than Gram-negative strains and extensive contact of the membrane with the immobilized

quaternary ammonium groups is less likely to occur, also under conditions of electrostatic

attraction. This might explain why the surface growth of Gram-positive bacteria is less affected

by the substratum charge. Several groups have also reported a reduction of viable counts, highest

for Gram-negative bacteria, when adding positively charged insoluble powders to bacterial

suspensions [21,22]. However, from these experiments is not clear whether this reduction is the

result of strong bacterial binding to the particles or of reduced viability of planktonic organisms.

Our results, however, clearly show that a positively charged surface can totally inhibit growth of

adhering bacteria.

Chapter 4

66

Initial bacterial adhesion has always been recognized as an essential step in biofilm

formation. This study shows that when a biomaterial surface is more negatively charged, this

may reduce the chance of bacterial adhesion, therewith delaying the formation of a biofilm.

Alternatively, positively charged surfaces are more adhesive, but the strong electrostatic

attraction of the organisms impedes surface growth of Gram-negative rods. By consequence, this

study points to a new pathway for the development of biomaterials with a low risk of infection

and complements current approaches based on preparing non-adhesive surfaces. Moreover, as

adhesion and growth appear to be oppositely affected by the surface characteristics of a

biomaterial, these results may explain why several biomaterials that have been found non-

adhesive under in vitro conditions, showed huge biofilm formation once implanted in the human

body. Everaert et al. [23], for instance, found that hydrophilized silicone rubber attracted lower

numbers of yeasts and bacteria under laboratory conditions in a parallel plate flow chamber than

authentic, hydrophobic silicone rubber, but during use as a voice prosthesis a much thicker

biofilm formed on the hydrophilized silicone rubber.

In conclusion, in order to develop biomaterial surfaces with a low risk of infection, in vitro

studies should not only account for initial adhesion, but should also include surface growth of the

adhering bacteria because adhering bacteria do not necessary grow well on all surfaces, as

indicated here for Gram-negative bacteria on positively charged surfaces.

References


Science 237, 1588-95.

2. Verkerke, G.J., Schraffordt-Koops, H., Veth, R.P., Van Horn, J.R., Postma, L. and Grootenboer,

H.J. (1997). First clinical experience with a noninvasively extendable endoprosthesis: a limb-saving

procedure in children suffering from a malignant bone tumor. Artificial Organs 21, 413-7.

3. Jucker, B.A., Harms, H., Hug, S.J. and Zehnder, A.B. (1997). Adsorption of bacterial surface

polysaccharides on mineral oxides is mediated by hydrogen bonds. Colloids and Surfaces B 9, 331-43.

4. Hogt, A.H., Dankert, J. and Feijen, J. (1986). Adhesion of coagulase-negative staphylococci to

methacrylate polymers and copolymers. Journal of Biomedical Materials Research 20, 533-45.

5. Harkes, G., Feijen, J. and Dankert, J. (1991). Adhesion of Escherichia coli on to a series of

poly(methacrylates) differing in charge and hydrophobicity. Biomaterials 12, 853-60.


67

6. Van der Mei, H.C., Brokke, P., Dankert, J., Feijen, J. and Busscher, H.J. (1992). Influence of

electrostatic interactions on the deposition efficiencies of coagulase-negative staphylococci to collector

surfaces in a parallel plate flow chamber. Journal of Dispersion Science and Technology 13, 447-58.

7. Van Loosdrecht, M.C.M., Lyklema, J., Norde, W. and Zehnder, A.J.B. (1990). Influence of interfaces

on microbial activity. Microbiological Reviews 54, 75-87.




surfaces. Journal of Biomaterials Science: Polymer Edition 3, 403-18.






12. Hogt, A.H., Gregonis, D.E., Andrade, J.D., Kim, S.W., Dankert, J. and Feijen, J. (1985). Wettability

and ζ potentials of a series of methacrylate polymers and copolymers. Journal of Colloid and Interface

Science 106, 289-98.

13. Mortimer, G.A. (1965). A new high-temperature free-radical source. Journal of Organic Chemistry 30,

1632-4.

14. Van Wagenen, R.A. and Andrade, J.D. (1980). Flat plate streaming potential investigations:

hydrodynamics and electrokinetic equivalency. Journal of Colloid and Interface Science 76, 305-14.

15. Hiemenz, P.C. (1977). Electrophoresis and other electrokinetic phenomena. In Principles of colloid and

surface chemistry. Edited by Lagowski, J.J. (Marcel Dekker, New York & Basel) p. 452-87.

16. Noordmans, J., Kempen, J. and Busscher, H.J. (1993). Automated image analysis to determine zeta

potential distributions in particulate microelectrophoresis. Journal of Colloid and Interface Science 156,

394-9.



18. Gottenbos, B., Van der Mei, H.C. and Busscher, H.J. (1999). Models for studying initial adhesion and

surface growth in biofilm formation on surfaces. Methods in Enzymology 310, 523-34.

19. Domagk, G. (1935). Eine neue Klasse von Desinfektionsmitteln. Deutsche Medizinische Wochenschrift 61,

829-31.

20. Lawrence, C.A. (1968). Quaternary ammonium surface-active disinfectants. In Disinfection, sterilization

and preservation. Edited by Lawrence, C.A. and Block, S.S. (Lea and Febiger, Philadelphia) p. 430-52.

21. Speier, J.L. and Malek, J.R. (1982). Destruction of microorganisms by contact with solid surfaces.

Journal of Colloid and Interface Science 89, 68-76.




Chapter 4

68

23. Everaert, E.P.J.M., Van de Belt-Gritter, B., Van der Mei, H.C., Busscher, H.J., Verkerke, G.J., Dijk,

F., Mahieu, H.F. and Reitsma, A. (1998). In vitro and in vivo microbial adhesion and growth on argon

plasma-treated silicone rubber voice prostheses. Journal of Materials Science: Materials in Medicine 9,

147-57.

69

Positively charged biomaterials exert antimicrobial

effects on Gram-negative bacilli in rats

Bart Gottenbos, Henny C. van der Mei, Flip Klatter, Dirk W. Grijpma, Jan Feijen, Paul

Nieuwenhuis and Henk J. Busscher

Biomaterial-centered infection is a much-dreaded complication associated with the use of

biomedical implants. Although positively charged biomaterial surfaces stimulate bacterial

adhesion, it has been suggested that surface growth of adhering Gram-negative bacilli is

inhibited on positively charged surfaces. In the present paper, we determined the infection rate of

differently charged poly(methacrylates) in rats. To this end, 2 x 106 cm-2 Escherichia coli O2K2

or 2 x 104 cm-2 Pseudomonas aeruginosa AK1 were seeded on glass discs coated with three

differently charged poly(methacrylates) coatings in a parallel plate flow chamber. Three rats

received six subcutaneous discs (two discs of each charge variant) seeded with E. coli, while

three other rats received discs seeded with P. aeruginosa. The numbers of viable bacteria on the

surfaces were determined 48 h after implantation. On 50% of all positively charged discs viable

E. coli were absent, while the negatively charged discs were all colonized by E. coli. P.

aeruginosa, however, were isolated from both positively and negatively charged discs. Probably,

P. aeruginosa can circumvent the antimicrobial effect of the positive charge through the

formation of extracellular polysacharides.

Submitted to Biomaterials

5

Chapter 5

70

Introduction

Infection is the most common cause of biomaterial implant failure in modern medicine [1,2].

Adhesion and subsequent surface growth of bacteria on biomedical implants and devices

causes the formation of a biofilm, in which the so-called “glycocalix” embeds the infecting

bacteria, and offers it protection against the host immune system and antibiotic treatment [3].

Recently, we found that adhering Escherichia coli O2K2 and Pseudomonas aeruginosa AK1

were unable to grow in vitro on poly(methacrylates) containing positively charged quaternary

ammonium groups [4,5]. Although bacterial adhesion was initially lower on negatively

charged poly(methacrylates), this effect was rapidly counterbalanced by surface-growth of the

adhering bacteria. It was concluded, that whether biomaterial-centered infections occur under

clinical conditions, depends on a critical balance between the effects of biomaterial surface

properties on bacterial adhesion and their inverse effects on growth.

The aim of this study was to determine, within a homologous series of three

methacrylate polymers and copolymers with varying surface charge, whether positively

charged surfaces have a lower risk of becoming infected under in vivo conditions. To this end,

E. coli O2K2 or P. aeruginosa AK1 were allowed to adhere on poly(methacrylates) with

different surface charges, after which the materials were implanted subcutaneously in rats,

and evaluated after 48 h for numbers of viable bacteria.


Animals

Six male, 14 week old, specific-pathogen free Albino Oxford rats weighing 300 g were used.

The animals were maintained under clean conventional conditions and fed standard rat chow

and water ad libitum. The animals were allowed to acclimatize to our laboratory conditions

for 2 weeks before experiments. All animals received humane care in compliance with the

“Principles of Laboratory Animal Care” (NIH Publication No.85-23, revised 1985) and the

Dutch Law on Experimental Animal Care.

Antimicrobial effects of biomaterials in rats

71

Bacteria

E. coli O2K2 was cultured in brain heart infusion in phosphate buffered saline (PBS) and P.

aeruginosa AK1 in nutrient broth in PBS. First, a strain was streaked and grown overnight at

37°C from a frozen stock on a blood agar plate. A colony was used to inoculate 5 ml of growth

medium, which was incubated at 37°C in ambient air for 24 h and used to inoculate a second

culture (100 ml) that was grown for 18 h. The bacteria from the second culture were harvested by

centrifugation and washed twice with sterile Millipore-Q water. Subsequently, the bacteria were

resuspended in PBS. The concentration of bacteria was determined using a counting chamber

and adjusted to 1 x 109 cells ml-1 in PBS.

Polymer synthesis

Polymers and polymer films were prepared as described previously [6]. Briefly, homopolymers

of methyl methacrylate (PMMA) and copolymers of MMA with either 15 mol% methacrylic

acid (PMMA/MAA 85/15) or 15 mol% trimethylaminoethyl methacrylate chloride

(PMMA/TMAEMA-Cl 85/15) were synthesized by radical polymerization of the monomers

using 2,2’-azobis(methyl isobutyrate) as an initiator. Glass discs (diameter 8 mm, 1 mm thick)

were cleaned by immersion in a mixture of hydrochloric acid (37%) and nitric acid (65%), ratio

3:1 (v/v) for 20 h. After extensive rinsing with double de-ionized water and ethanol, the discs

were dried in vacuo at 60 °C for 3 h. The discs were silanized with n-propyltrimethoxysilane in

case of PMMA and with γ-aminopropyltriethoxysilane in case of PMMA/MAA. The polymer

films were prepared on both sides of the discs by spin coating [6]. Finally, the polymer films

were dried in vacuo at 60 °C for 18 h in sterile petridishes to remove any solvent remnants. The

zeta potentials, indicative of the surface charge, of the polymer films in PBS were determined as

described previously [7], and were –18 mV for PMMA/MAA, -12 mV for PMMA and +12 mV

for PMMA/TMAEMA-Cl.

Seeding of bacteria and implantation.

The bacteria were seeded on the coated glass discs in a parallel plate flow chamber, according

to a procedure that has previously been described in detail [8]. A bacterial suspension of 1 x

109 cells ml-1 in PBS was passed through the chamber at a flow rate of 0.025 ml s-1 for 1 h.

Then, flow was switched to PBS without bacteria to remove unbound organisms from the

tubes and the flow chamber under the same flow rate for 30 min. Phase-contrast microscope

Chapter 5

72

images were taken of each polymer coating, and the total numbers of adhering bacteria were

counted using image analysis software. Finally, the tubes were blocked and the chambers

were opened. The discs were carefully removed and either put in 5 ml reduced transport fluid

(RTF) for CFU determination or subcutaneous implantation in rats.

Before implantation, the backs of the rats were shaved and disinfected with

chlorhexidine in 70% ethanol after induction of inhalation anesthesia with N2O/O2 (3/2) and

halothane. Six 1 cm incisions were made, three on either side of the middle line, at least 2 cm

apart. Subcutaneous pockets of at least 2 cm deep were created. At either side of the middle

line three differently charged discs were inserted and the incisions were closed with 2 stitches.

Three rats received discs seeded with E. coli and three others received discs seeded with P.

aeruginosa. After 48 h the rats were terminated with CO2. After skin disinfection and

opening, the discs were carefully removed and added to 5 ml RTF for enumeration.

For enumeration, discs in RTF were first treated in a sonicating water bath for 10 min

to remove adhering bacteria. Removal of the bacteria was verified with phase contrast

microscopy. Then, suspensions were diluted and streaked on blood agar to determine the

number of colony forming units (CFU). The lowest number that could be detected was 50

bacteria per disc. The infection incidence was defined as the percentage of explanted discs per

charge variant from which bacteria could be harvested.

Table 1. Infection incidence (number of infected discs over the total number of discs involved) for a

homologous series of subcutaneously implanted differently charged poly(methacrylates) and CFU ± standard

errors (in 104 cm-2) before and after 48 h of subcutaneous implantation in rats (two discs of each charge variant

were implanted in three rats). Note that the CFU presented are averages, only pertinent to the infected discs

Bacterial strain Parameter PMMA/MAA

(-- charge)

PMMA

(- charge)

PMMA/TMAEMA-Cl

(+ charge)

E. coli O2K2 Incidence 6/6 6/6 3/6

CFU before 520 ± 280 960 ± 280 1440 ± 560

CFU after 1.1 ± 0.3 70 ± 60 4.8 ± 1.9

P. aeruginosa AK1 Incidence 5/6 6/6 6/6

CFU before 5.2 ± 1.2 9.2 ± 0.8 7.6 ± 1.2

CFU after 5.2 ± 4.1 2.5 ± 1.1 36 ± 26


73

Results and discussion

Table 1 shows the infection incidence and the mean numbers of CFU on the infected materials

before and after 48 h of implantation. The infection incidence of E. coli O2K2 was twice as

low on the positively charged PMMA/TMAEMA-Cl than on the negatively charged surfaces

as bacteria were recovered from only three discs. Moreover, the number of viable E. coli

decreased significantly (Student t-test, p<0.05) during implantation, especially on

PMMA/MAA and PMMA/TMAEMA-Cl. On the negatively charged PMMA/MAA and

PMMA, the numbers of CFU for P. aeruginosa AK1 were similar before and after

implantation. Unexpectedly, the number of CFU on positively charged PMMA/TMAEMA-Cl

increased during 48 h of implantation, although this increase was not statistically significant

(Student t-test, p>0.05). Note that no viable P. aeruginosa were isolated from one

PMMA/MAA disc. Also the CFU on the discs before implantation were not significantly

different between the three differently charged discs, although the total number of adhering

bacteria (viable and non-viable) as determined by phase contrast microscopy was significantly

higher on the positively charged discs (compared to PMMA 4 times for E. coli and 2 times for

P. aeruginosa).

From these results, it can be concluded that positively charged poly(methacrylate)

surfaces have a lower risk of becoming infected by E. coli O2K2 under in vivo conditions than

negatively charged surfaces. As before implantation the numbers of viable E. coli were

similar on all charged surfaces, this is probably due to physical inhibition of bacterial surface

growth, as also observed in vitro [5]. Although in vitro inhibition of surface growth was also

seen for P. aeruginosa AK1 on positively charged surfaces [4], results for this organism

evidently do not hold in vivo. Possibly, P. aeruginosa AK1 produces considerably larger

amounts of extracellular polysacharides than E. coli O2K2, creating a barrier between its cell

surface and the antimicrobial positively charged surface. Extracellular polysacharides are

known virulence factors in bacterial infection, protecting the organisms against the hosts

immune system [9], which explains why P. aeruginosa AK1 can grow on the implanted discs,

whereas E. coli were readily eradicated by the immune system of the rats, especially on the

positively charged discs.

In conclusion, positively charged biomaterials with reduced surface growth of

adhering bacteria have a lower risk of infection in vivo than negatively charged surfaces. The

development of such antimicrobial surfaces may have several applications, for instance on

Chapter 5

74

medical devices used in the urogenital tract, where catheter-associated infections are rampant

[10,11] and E. coli constitute the most important causative organisms for infection [12].

Especially now that many antibiotics have become ineffective due to bacterial resistance

[13,14], the application of positively charged surfaces may provide a clinical solution to

several types of biomaterial-centered infections.

References


Science 237, 1588-95.






4. Gottenbos, B., Grijpma, D.W., Van der Mei, H.C., Feijen, J. and Busscher, H.J. (1999). Initial

adhesion and surface growth of Pseudomonas aeruginosa on negatively and positively charged

poly(methacrylates). Journal of Materials Science: Materials in Medicine 10, 853-5.













10. Nickel, J.C., Costerton, J.W., McLean, R.J. and Olson, M. (1994). Bacterial biofilms: influence on the

pathogenesis, diagnosis and treatment of urinary tract infections. Journal of Antimicrobial Chemotherapy,

supplement A 33, 31-41.

11. Habash, M. and Reid, G. (1999). Microbial biofilms: their development and significance for medical

device-related infections. Journal of Clinical Pharmacology 39, 887-98.

12. Denstedt, J.D., Wollin, T.A. and Reid, G. (1998). Biomaterials used in urology: current issues of

biocompatibility, infection, and encrustation. Journal of Endourology 12, 493-500.


75

13. Gould, I.M. (1999). A review of the role of antibiotic policies in the control of antibiotic resistance.

Journal of Antimicrobial Chemotherapy 43, 459-65.

14. Walsh, C. (2000). Molecular mechanisms that confer antibacterial drug resistance. Nature 406, 775-81.

77

Late hematogenous infection of subcutaneous

implants in ratsBart Gottenbos, Flip Klatter, Henny. C. van der Mei, Henk J. Busscher and Paul Nieuwenhuis

Late biomaterial-centered infection is a major complication associated with the use of

biomaterial implants. In this study subcutaneously implanted biomaterials in rats were

hematogenously challenged with bacteria 4 weeks after implantation. Bacteria were spread by

either intravenous injection or stimulation of bacterial translocation. It was found that none of

the biomaterials was infected by hematogenous spread, whereas 5% of the implants were

infected by perioperative contamination. We conclude that late hematogenous infection of

subcutaneous biomaterials does not occur in the rat. Also in man, there are growing doubts

whether implants actually become infected through hematogenous routes or whether late

infections are caused by delayed appearance of perioperatively introduced bacteria.

Submitted to Clinical and Diagnostic Laboratory Immunology

6

Chapter 6

78

Introduction

A severe complication associated with the use of biomaterial implants is failure due to

infection. About half of all biomaterial-centered infections occur months to even years after

deep tissue implantation. Controversy exists concerning the origin of the infecting

microorganisms in these late infections. Either bacteria spread hematogenously from

endogenous foci are inserted during implantation and stay clinically unnoticed for a long time,

designated as delayed infections [1].

Most hematogenous infections are believed to arise from infected skin lesions

producing relapsing bacteremia [2]. This is supported by the fact that in most (56%)

infections, where hematogenous spreading is suspected, staphylococci are involved, which are

part of the normal skin flora. Also dental or other surgical interventions, bacteriuria, intestinal

surgery or pneumonia have been proposed as possible causes of hematogenous spreading of

bacteria. Streptococci are found in 15% of all late biomaterial-centered infections, while

common intestinal bacteria, for example Escherichia coli and Pseudomonas aeruginosa, are

responsible for 23% of the infections [1,2]. Another possible mechanism for hematogenous

spreading from the intestinal tract is bacterial translocation [3], i.e. the escape of mainly

Gram-negative rods through the intestinal wall [4].

Bacterial translocation (BT) can be promoted by nutritional factors, like total

parenteral nutrition, fluid elemental nutrition, protein malnutrition [5] and vitamin A

deficiency [6], hemorrhagic shock, extensive thermal injury and endotoxin [7]. Interestingly,

also intraperitoneal implants promote BT [8,9].

In animal studies on biomaterial-centered infections, man-derived bacteria are

frequently used. In man, however, biomaterial-centered infections are caused in most cases by

their own commensal microflora toward which the immune system is more tolerant than to

foreign flora [10,11]. As tolerated microorganisms probably survive longer in the circulatory

system, it can be expected that their chance to cause biomaterial-centered infections is larger

as compared with non-immunotolerated microorganisms.

The aim of this study was to determine whether hematogenous spreading of bacteria,

after healing of the implantation wound, infects subcutaneous (s.c.) implants in rats. To this

end, rats were intravenously (i.v.) injected with Staphylococcus aureus, Staphylococcus

epidermidis or P. aeruginosa or their own total fecal flora, 4 weeks after implantation of a

Late hematogenous biomaterial-centered infections

79

biomaterial. To investigate the possibility of infection with translocating intestinal bacteria,

BT was promoted either through special diets or through an intraperitoneal implant.


Rats

Forty-eight male, 12 weeks old, specific-pathogen free Albino Oxford rats weighing 220-260

g were used. The animals were housed in a standard temperature-controlled environment

(22ºC), in macrolon cages and kept on a 12 h light/dark cycle. The rats were fed normal rat

chow, unless otherwise stated, and had sterile tap water supplied ad libitum. Animals were

allowed to acclimatize to our laboratory conditions for 2 weeks before the experiments. All

animals received humane care in compliance with the “Principles of Laboratory Animal Care”

(NIH Publication No.85-23, revised 1985) and the Dutch Law on Experimental Animal Care.

Bacteria

Man-derived S. aureus ATCC 12600 and S. epidermidis HBH2 102 were cultured in tryptone

soya broth (OXOID, Basingstoke, UK) in phosphate buffered saline (PBS) and man-derived P.

aeruginosa AK1 in nutrient broth (OXOID) in PBS. First, a strain was streaked and grown

overnight at 37°C from a frozen stock on a blood agar plate. A colony was used to inoculate 5 ml

of growth medium, which was incubated at 37°C in ambient air for 24 h and used to inoculate a

second culture (150 ml) that was grown for 18 h. The bacteria from the second culture were

harvested by centrifugation (5 min, 5000 g for staphylococci and 10,000 g for P. aeruginosa) and

washed twice with sterile Millipore-Q water. Subsequently, the bacteria were resuspended in

sterile 0.9% NaCl solution, and S. epidermidis was sonicated on ice to disrupt aggregates.

Gut bacteria were harvested from fresh feces of the rats. The feces (2-3) were

suspended in 10 ml anaerobic 0.9% NaCl solution. The suspension was centrifuged for 2 min

at 250 g to remove larger particles. Supernatants were centrifuged at 10,000 g for 20 min to

spin down the bacteria. Finally, the pellets were suspended in 10 ml 0.9% NaCl. The fecal

flora contained anaerobic bacteria (70%), E. coli (20%), lactobacilli (7%) and streptococci

(3%).

Chapter 6

80

Biomaterials

Discs (diameter 8 mm, 0.5 mm thick) without sharp edges were made of commercially

available silicone rubber (SR), polyethylene (PE), polypropylene (PP),

poly(tetrafluoroethylene) (PTFE), poly(ethylene terephthalate) (PET), poly(methyl

methacrylate) (PMMA), polyurethane (PU) (pellethane 2363-75D) and glass. The discs were

cleaned in a 2% RBS 25 detergent solution under simultaneous sonication and thoroughly

rinsed in demineralized water, sterilized in 70% ethanol and finally washed with sterile

Millipore-Q water.

Implantation

Each rat received only four subcutaneous biomaterial discs as space was limited. After

induction of inhalation anesthesia with N2O/O2 (3/2) and halothane, the backs of the rats were

shaved and disinfected with 0.5% chlorhexidine in 70% ethanol. Four 1 cm incisions were

made, two on either side of the middle line, at least 3 cm apart. Subcutaneous pockets of at

least 2 cm deep were created. The 4 different implants were inserted as deep as possible. The

incision was then closed with degradable suture material. The surgical instruments used were

disinfected after each surgical action.

Figure 1. Design of intravenous injection experiments, involving a total of 28 rats, for practical reasons

subsequently divided in 5 groups, depending on the infecting organisms. Number of rats per group are indicated

between brackets.

Intravenous inoculation experiment

Figure 1 shows the experimental design used for i.v. inoculation. Half of the 28 rats received a

SR, PTFE, PP and PE disc while the other half received a PU, PET, PMMA and glass disc.

Subcutaneous implantation (28)

InjectionS. aureus (8)

S. epidermidis (8)

InjectionOwn fecal flora (3)

Foreign fecal flora (1)

4 weeks

4 weeks1 week

Sacrifice

P. aeruginosa (8)


81

After 4 weeks, 0.5 ml of bacterial suspension was injected in the tail vein. From each

biomaterials group 12 rats were injected with a different one of the following 12 bacterial

suspensions: 3 x 107; 108; 3 x 108 or 109 CFU ml-1 S. aureus or S. epidermidis or 108; 3 x 108;

109 or 3 x 109 CFU ml-1 P. aeruginosa. Three other rats received suspensions with 3 x 108; 109

or 3 x 109 CFU ml-1 of their own fecal flora, while the last one (i.e. 3 x 109 CFU ml-1 fecal

flora) was also injected in one other rat, thus receiving foreign fecal flora.

Stimulated bacterial translocation experiment

Figure 2 shows the experimental design used to stimulate bacterial translocation.

Figure 2. Design of bacterial translocation experiments, involving a total of 20 rats, divided in four groups: rats

on vitamin A deficient diet (-vit A), rats on liquid diet and rats with an intraperitoneal implant with or without

vitamin A deficiency. Number of rats per group are indicated between brackets.

The rats in this experiment all received a disc of SR, PTFE, PET and PMMA. 5 rats were fed

normal rat chow, while the other 15 were fed vitamin A free rat chow (Hope Farms, The

Netherlands), starting directly after s.c. implantation, as vitamin A deficiency has been

reported to occur 4 weeks after the onset of the diet [6]. 4 weeks after implantation the diet of

5 vitamin A deficient rats was changed to total liquid elemental nutrition (Nutrison powder,

Nutricia, Zoetermeer, The Netherlands), which contained vitamin A, made according to the

manufacturers descriptions with sterile demineralized water in sterile drinking bottles. Also 4

weeks after s.c. implantation a proteograft patch (dimensions 3.3 x 3.3 cm, similar to Dacron

velour material, Braun, Oss, The Netherlands) was intraperitoneally implanted in 5 vitamin A

deficient rats and in the 5 rats on normal rat chow. To this end the rats were anesthetized with

Subcutaneous implantation (20)

Intraperitoneal implant

+Vit A (5)-Vit A (5)

Sacrifice

4 weeks4 weeks

4 weeks

Food induced

-Vit A (5) Liquid diet (5)

Chapter 6

82

N2O/O2 (3/2) and halothane and their anterior side was shaved and disinfected. A 5 cm

incision was made in the skin longitudinally over the middle line. Then a 4 cm incision was

made in the abdominal wall. The implant was inserted close to the gut taking care it was not

irritating the bladder or the liver. The abdominal wall was closed and the skin was closed

separately. The rats received postoperatively pain killers (0.1 mg kg-1 Temgesic a day) for 1

week.

Harvesting

After induction of anesthesia with N2O/O2 (3/2) and halothane, the back of the rats was

shaved and disinfected. The subcutaneous implants were explanted and stored in 5 ml sterile

reduced transport fluid (RTF). Swabs were taken of the pockets and streaked onto blood agar.

Subsequently, the anterior side of the rats was shaved and disinfected. The abdominal cavity

and chest were opened through a midline incision, and 0.1 ml of ventricular blood, a swab of

the inside of the abdominal wall, the intraperitoneal implant when appropriate, a halved

kidney, the halved spleen and a section of the liver were streaked onto blood agar plates. In

the BT experiment also a section of the lungs was taken and streaked onto blood agar, and the

mesenteric lymph nodes (MLN) were harvested and homogenized in 5 ml RTF. The rats were

terminated by a cut in the heart. The MLN suspension and the biomaterials in RTF were

sonicated on ice for 5 min to remove the attached bacteria, and cultured on blood agar. The

blood agar plates were incubated aerobically at 37ºC. The subcutaneous implants from the

rats injected with fecal microflora and from the rats in the BT experiment, including the

intraperitoneal implant, were also anaerobically cultured. Also the MLN were anaerobically

cultured. The plated samples were considered infected if more than 2 of the same colonies

were found on the agar plate (corresponding to more than 100 CFU per biomaterial disc).

Bacteria harvested were characterized by colony morphology and Gram-staining.

Results

Intravenous inoculation experiment

The two rats that had received the highest dose of S. aureus and the rat recieving the highest

dose of its own fecal flora, were terminated within 4 days because of severe illness and

excluded from the study. The rat receiving foreign fecal flora died after 17 days due to sepsis,

but was not excluded from the study.


83

Table 1 shows the numbers of positive organs and biomaterials cultures, together with

the number of CFU isolated from the discs. Out of the 100 implanted discs 92 showed no

infections. Moreover, most of the infected biomaterial discs revealed different bacterial strains

than used for injection. These were staphylococci on both PE discs, two staphylococcal

strains on one PU disc and a staphylococcal and streptococcal strain on the other PU disc. The

SR disc was infected with two different strains of Gram-positive rods.

Table 1. Biomaterial-centered infections in intravenous injection and bacterial translocation model, including

culture positive organs and different s.c. implanted biomaterials found infected and the numbers of CFU per

biomaterial disc. Numbers between brackets present: (number of positive organs or discs/number of animals

involved). Note that 168 out of the 180 implanted biomaterial discs were sterile.

Intravenous injection model

Bacteria Positive organs Positive biomaterial discs

Type CFU (x103)

S. epidermidis HBH2102 Kidney (8/8) PUa (1/4) 8.2

S. aureus ATCC 12600 Kidney (8/8) PUa (1/4); SRa (1/4) 1.4; 29

P. aeruginosa AK1 None PEa (2/4) 100; 100

Own fecal flora None None

Foreign fecal flora All organs SR (1/1); PP (1/1); PE (1/1) 15; 2.5; 1.0

Bacterial translocation model

Animal condition Positive organs Positive biomaterial discs

Type CFU (x103)

Vitamin A deficiency None None

Liquid diet None SRb (1/5); PMMAb (1/5) 0.45; 100

Intraperitoneal implant

and vitamin A deficiency

MLN (3/5); Kidney (1/5);

Peritoneum (1/5);

Intraperitoneal implant (1/5)

None

Intraperitoneal implant MLN (4/5); Kidney (1/5);

Liver (1/5); Spleen (1/5);

Blood (1/5); Peritoneum (1/5)

PETb (1/5)

PTFEb (1/5)

4.5

10

a Colony form did not correspond to the injected strains.b Colony form and Gram-staining did not correspond with translocated bacteria.

Stimulated bacterial translocation experiment

The results for the BT experiment are also shown in Table 1. Bacterial translocation to the

MLN and some organs was only observed in the intraperitoneal implant group and found to

Chapter 6

84

be due to Gram-negative bacilli (E. coli) and Gram-positive branched rods (Actinomyces

spp.). Out of the 80 implanted discs 76 showed no infections. Moreover, the translocating

species were not found on the infected biomaterial discs. A PET disc showed two different

strains of staphylococci, while the other infected discs revealed one staphylococcal strain.

Discussion

In this study, the susceptibility of subcutaneously implanted biomaterials to late

hematogenous infection was determined in rats, 4 weeks after biomaterial implantation. Two

routes of hematogenous spreading of bacteria were used: either single i.v. injection of bacteria

as a model for transient bacteremia, or by promoted bacterial translocation from the digestive

tract. None of the respective biomaterial discs in the non-septic rats became infected by i.v.

injected bacterial strains. Five biomaterial discs revealed bacteria that were probably

originating from perioperative contamination. The 4 infected biomaterial discs in the

increased BT group showed exclusively staphylococcal strains, suggesting that translocation

from the intestinal tract was not the source of infection, since staphylococci are numerous on

the skin, but scarcely found in the gut flora [12]. Furthermore no staphylococci were found in

the MLN, indicating absence of staphylococcal translocation. Most probably, these bacteria

also originated from perioperative contamination. Perioperative contaminations are likely to

occur in animal experiments as usually no ultra-clean operating rooms are used and antibiotic

prophylaxis is not common [13,14]. All infected biomaterials, were relatively hydrophobic,

while glass, a hydrophilic material, was not involved in any biomaterial-centered infection.

This corresponds with earlier in vitro findings that surface growth of staphylococci is slow on

glass as compared with the other materials [15].

To our knowledge, the use of late hematogenous infection models with

subcutaneously implanted biomaterials has not been reported in rats before. In mice [16], i.v.

injection of 1 x 107 S. aureus did not yield infection of subcutaneously implanted

biomaterials, 1 month after implantation, while a higher dose 1 x 108 S. aureus killed the

mice. However, in rabbits, 6 to 8 weeks after total joint replacement, Blomgren and Lindgren

[17] successfully induced hematogenous infections in 40% of the cases through i.v. injection

of around 109 S. aureus (note this is a 10 to 100 fold higher dose than used in our study).

Hematogenous infection directly after implantation yielded infection in 80% of the animals

[18]. Also Southwood et al. [19] showed a similar decrease in hematogenous infection rate in


85

rabbits from 40 % immediately after surgery to 10% after 3 weeks of implantation. Vascular

grafts in dogs were infected by i.v. challenge of 108 S. aureus 3 to 6 months after

implantation, yielding infections in 10% to 80% of the animals, depending on the type of graft

[20]. Interestingly, a similar study in rats revealed that the infection rate of caval vein grafts

was reduced from 100% to zero during 2 weeks of implantation as a result of increased

endothelialization, while the infection rate of aorta grafts was still 100% after these 2 weeks

[21].

Evidently, hematogenous biomaterial-centered infections can be induced directly after

implantation, but are much more difficult to achieve after prolonged implantation time.

Essentially, the occurrence of biomaterial-centered infections is a race for the surface [22]

between infecting microorganisms and host cells. When infecting organisms arrive long-time

after implantation on a biomaterial surface, the race is won in most cases by host cells and the

biomaterial surface is out of reach for adhering organisms. Yet, many late biomaterial-

centered infections in man, most notably those associated with orthopedic implants are said to

be hematogenous in origin [1,2,14,23], although it is also suggested that the hematogenous

route of infection will only occur in immuno-compromised patients. For example, 40% to

100% of all late hematogenous orthopedic implant infections are found in patients with

rheumatic arthritis [24] using immuno-modulating drugs. Late infections associated with

dental procedures have been reported mostly in diseased patients with drug or irradiation

induced immuno-suppression, insulin dependent diabetes mellitis or hemophilia [25]. At this

point it must be noted that in clinical practice infection is often assumed to be of

hematogenous origin without attempts to obtain any proof, for instance, by culturing blood or

joint fluids [1]. As many strains, including S. aureus and Gram-negative rods, can survive

intracellularly in epithelial and scar tissues, therewith circumventing the host’s immune

system for prolonged periods of time [26], it is suggested more and more that many

biomaterial-centered infections assumed to be of hematogenous origin actually results from

delayed appearance of perioperatively introduced bacteria. These suggestions would be in line

with the results of this study, demonstrating that it is virtually impossible to create a

biomaterial-centered infection in rats by the hematogenous route, although of course

differences between the immune system of rat and man may be of crucial importance here.

Chapter 6

86

References

1. Ahlberg, A., Carlsson, A.S. and Lindberg, L. (1978). Hematogenous infection in total joint replacement.

Clinical Orthopaedics and Related Research 137, 69-75.

2. Bengtson, S., Blomgren, G., Knutson, K., Wigren, A. and Lidgren, L. (1987). Hematogenous infection

after knee arthroplasty. Acta Orthopaedica Scandinavica 58, 529-34.

3. Van Leeuwen, P.A., Boermeester, M.A., Houdijk, A.P., Ferwerda, C.C., Cuesta, M.A., Meyer, S. and

Wesdorp, R.I. (1994). Clinical significance of translocation. Gut 35, S28-34.

4. O'Boyle, C.J., MacFie, J., Mitchell, C.J., Johnstone, D., Sagar, P.M. and Sedman, P.C. (1998).

Microbiology of bacterial translocation in humans. Gut 42, 29-35.

5. Deitch, E.A. (1994). Bacterial translocation: the influence of dietary variables. Gut 35, S23-7.

6. Wiedermann, U., Hanson, L.A., Bremell, T., Kahu, H. and Dahlgren, U.I. (1995). Increased

translocation of Escherichia coli and development of arthritis in vitamin A-deficient rats. Infection and

Immunity 63, 3062-8.

7. Edmiston, C.E. and Condon, R.E. (1991). Bacterial translocation. Surgery, Gynecology and Obstetry 173,

73-83.

8. Mora, E.M., Cardona, M.A. and Simmons, R.L. (1991). Enteric bacteria and ingested inert particles

translocate to intraperitoneal prosthetic materials. Archives of Surgery 126, 157-63.

9. Guo, W., Andersson, R., Ljungh, A., Wang, X.D. and Bengmark, S. (1993). Enteric bacterial

translocation after intraperitoneal implantation of rubber drain pieces. Scandinavian Journal of

Gastroenterolology 28, 393-400.

10. Berg, R.D. and Savage, D.C. (1975). Immune responses of specific pathogen-free and gnotobiotic mice to

antigens of indigenous and nonindigenous microorganisms. Infection and Immunity 11, 320-9.








14. Blackburn, W.D. and Alarcon, G.S. (1991). Prosthetic joint infections. A role for prophylaxis. Arthritis

and Rheumatism 34, 110-7.




16. Merritt, K., Shafer, J.W. and Brown, S.A. (1979). Implant site infection rates with porous and dense

materials. Journal of Biomedical Materials Research 13, 101-8.

17. Blomgren, G. and Lindgren, U. (1981). Late hematogenous infection in total joint replacement: studies of

gentamicin and bone cement in the rabbit. Clinical Orthopaedics and Related Research 155, 244-8.


87

18. Blomgren, G. and Lindgren, U. (1980). The susceptibility of total joint replacement to hematogenous

infection in the early postoperative period: an experimental study in the rabbit. Clinical Orthopaedics and

Related Research 151, 308-12.

19. Southwood, R.T., Rice, J.L., McDonald, P.J., Hakendorf, P.H. and Rozenbilds, M.A. (1985). Infection

in experimental hip arthroplasties. The Journal of Bone and Joint Surgery, British volume 67, 229-31.

20. Moore, W.S., Malone, J.M. and Keown, K. (1980). Prosthetic arterial graft material. Influence on

neointimal healing and bacteremic infectibility. Archives of Surgery 115, 1379-83.

21. Zdanowski, Z., Hallberg, E., Schalen, C. and Ribbe, E. (1994). Reduced susceptibility of

polytetrafluoroethylene vascular prostheses to colonization by Staphylococcus aureus following in situ

endothelialization. Artificial Organs 18, 448-53.


Science 237, 1588-95.

23. Maniloff, G., Greenwald, R., Laskin, R. and Singer, C. (1987). Delayed postbacteremic prosthetic joint

infection. Clinical Orthopaedics and Related Research 223, 194-7.

24. Poss, R., Thornhill, T.S., Ewald, F.C., Thomas, W.H., Batte, N.J. and Sledge, C.B. (1984). Factors

influencing the incidence and outcome of infection following total joint arthroplasty. Clinical Orthopaedics

and Related Research 182, 117-26.

25. LaPorte, D.M., Waldman, B.J., Mont, M.A. and Hungerford, D.S. (1999). Infections associated with

dental procedures in total hip arthroplasty. The Journal of Bone and Joint Surgery, British volume 81, 56-9.

26. Smith, R.P., Baltch, A.L., Franke, M.A., Michelsen, P.B. and Bopp, L.H. (2000). Levofloxacin

penetrates human monocytes and enhances intracellular killing of Staphylococcus aureus and Pseudomonas

aeruginosa . Journal of Antimicrobial Chemotherapy 45, 483-8.

89

In vitro and in vivo antimicrobial activity of

covalently coupled quaternary ammonium silane

coatings on silicone rubberBart Gottenbos, Henny C. van der Mei, Flip Klatter, Paul Nieuwenhuis and Henk J. Busscher

Biomaterial-centered infection is a dreaded complication associated with the use of biomedical

implants. In this paper, the antimicrobial activity of silicone rubber with a covalently coupled 3-

(trimethoxysilyl)-propyldimethyloctadecylammonium chloride (QAS) coating was studied in

vitro and in vivo. Gram-positive Staphylococcus aureus ATCC 12600, Staphylococcus

epidermidis HBH2 102, and Gram-negative Escherichia coli O2K2 and Pseudomonas

aeruginosa AK1 were seeded on the silicone rubber with and without QAS-coating, in the

absence or presence of adsorbed human plasma proteins. The viability of the adherent bacteria

was determined using a live/dead fluorescent stain and a confocal laser scanning microscope.

The coating reduced the viability of adherent staphylococci from 90% to 0%, and of Gram-

negative bacteria from 90% to 25%, while the presence of adsorbed plasma proteins had little

influence. The biomaterials were also subcutaneously implanted in rats for 3 or 7 days, while

preoperatively or postoperatively seeded with S. aureus ATCC 12600. Preoperative seeding

resulted in infection of 7 out of 8 silicone rubber implants against 1 out of 8 QAS-coated

silicone rubber implants. Postoperative seeding resulted in similar infection incidences on

both implant types, but the numbers of adhering bacteria were 70% lower on QAS-coated

silicone rubber. In conclusion, QAS-coated silicone rubber shows antimicrobial properties

against adhering bacteria, both in vitro and in vivo.

Submitted to Biomaterials

7

Chapter 7

90

Introduction

Infection is the most common cause of biomaterial implant failure in modern medicine [1,2].

Adhesion and subsequent surface growth of bacteria on biomedical implants and devices

causes the formation of a biofilm, in which the so-called “glycocalix” embeds the infecting

bacteria offering protection against the host immune system and antibiotic treatment [3].

Gram-positive Staphylococcus aureus, and Staphylococcus epidermidis are the predominant

infecting organisms, followed by Gram-negative bacilli like Escherichia coli and

Pseudomonas aeruginosa [4,5]. A possible approach to prevent biomaterial-centered

infections is to render the biomaterial surface antimicrobial properties by functionalization

with quaternary ammonium groups, which are widely known as disinfectants [6]. In this

approach no antimicrobial agents are leaching from the surface, providing long term

protection against bacterial colonization, and reducing the risk of developing antimicrobial

resistant microbial strains, as the concentration of antimicrobial groups is constantly above the

minimal inhibitory concentration [7]. As quaternary ammonium functionalized surfaces have

a high positive surface charge, they exert a strong adhesive force on negatively charged

bacteria, which has been proposed to physically inhibit surface growth of rod-shaped bacteria

[8].

Poly(methacrylates) with methyl or ethyl quaternary ammonium chloride side groups

showed antimicrobial activity [9,10] toward Gram-negative strains, although Gram-positive

staphylococci were little affected by these polymers. The antimicrobial effects of soluble

quaternary ammonium compounds increase with the length of the alkyl moieties on the

nitrogen atom [11-13], with an optimum chain lengths of 16 to 18 carbon atoms [14]. Less

soluble dendrimers functionalized with dimethyldodecylammonium chloride groups were also

found to have a strong antimicrobial effect towards S. aureus [15]. Cotton-polyester fabric

treated with 3-(trimethoxysilyl)-propyldimethyloctadecylammonium chloride (QAS) showed

significant antimicrobial activity towards most Gram-positive cocci [16]. QAS possesses a

silyl group, which can be covalently bound to glass and cotton, but also to many other

frequently employed biomaterials, including silicone rubber, after activation of the surface

with gas plasma [17,18].

The aim of this study was to determine the antimicrobial activity of a QAS coating on

silicone rubber in vitro toward two Gram-positive and two Gram-negative strains.

Subsequently, QAS coatings have been evaluated in vivo toward a S. aureus strain.

Antimicrobial activity of quaternary ammonium silane coatings

91


Animals

Eight male, 12 weeks old, specific-pathogen free Albino Oxford rats (Harlan Nederland,

Horst, The Netherlands) weighing 290 ± 30 g were used. The animals were maintained under

clean conventional conditions and fed standard rat chow and water ad libitum. The animals

were allowed to acclimatize to our laboratory conditions for 1 week before experiments. All

animals received humane care in compliance with the “Principles of Laboratory Animal Care”

(NIH Publication No.85-23, revised 1985) and the Dutch Law on Experimental Animal Care.

Bacterial strains

S. aureus ATCC 12600 and S. epidermidis HBH2 102 were cultured in tryptone soya broth

(OXOID, Basingstoke, UK) in phosphate buffered saline (PBS), E. coli O2K2 in brain heart

infusion (OXOID) in PBS and P. aeruginosa AK1 in nutrient broth (OXOID) in PBS. First, a

strain was streaked and grown overnight at 37°C from a frozen stock on a blood agar plate. A

colony was used to inoculate 5 ml of growth medium, which was incubated at 37°C in ambient

air for 24 h, and used to inoculate a second culture (150 ml for staphylococci and 100 ml for the

other bacteria) that was grown for 17 h. The bacteria from the second culture were harvested by

centrifugation (5 min, 5000 g or 10,000 g for P. aeruginosa) and washed twice with sterile

Millipore-Q water. Subsequently, the bacteria were resuspended in PBS. The S. epidermidis

suspension was sonicated on ice (3 x 10 s) to disrupt aggregates. The concentration of bacteria

was determined using a counting chamber and adjusted to the appropriate concentration in PBS.

For the in vivo experiment also the number of colony forming units (CFU) in suspension was

determined.

Silanization of silicone rubber and surface characterization

Discs with a diameter 8 mm were cut out of implant grade silicone rubber (SR) (Medin,

Groningen, The Netherlands) sheets of 0.5 mm thick. The discs were cleaned in a 2% RBS 35

detergent solution (Omniclean, Breda, The Netherlands) under simultaneous sonication and

thoroughly rinsed in demineralized water, sterilized in 70% ethanol, washed with sterile

Millipore-Q water and dried overnight at 80 °C in petridishes. For silanization (for reaction

scheme, see Figure 1), discs were oxidized in a glow-discharge reactor (a DC modified Edwards

Chapter 7

92

sputter coater S150B). Argon plasma treatment was done under 5 mbar argon pressure, at a

power of 7 W for 5 min, followed by exposure to ambient air. Subsequently, each oxidized

silicone rubber disc was covered with 40 µl 0.5% 3-(trimethoxysilyl)-

propyldimethyloctadecylammonium chloride (QAS, Dow Corning Corporation, USA) in

Millipore water. Coated silicone rubber discs were allowed to react and dry at 80 °C for 20 h.

For in vivo experiments, discs were coated on both sides. Sheets of SR (25 x 76 mm) were

similarly treated for streaming potential measurements.

Figure 1. Reaction scheme of silanization of silicone rubber. Silicone rubber was first oxidized by argon plasma

treatment, and subsequently reacted with 3−(trimethoxysilyl)propyldimethyloctadecylammonium chloride

(QAS) in water. Note that the QAS molecules were also able to react with itself, forming polymers.

For chemical and physical characterization, the quaternary ammonium silanized SR

(QAS-coated SR) was washed for 30 min in PBS followed by rinsing with demineralized water.

The chemical composition of SR and QAS-coated SR surfaces was determined by X-ray

CH3

CH3

CH3

CH3

CH3

Ar plasma

CH3

OH

CH3

CH3

OH

Si CH2 CH2 CH2 N+

CH3

C18H37

CH3

OCH3

CH3O

OCH3

H2O curing80 °C, 20 h

Silicone rubber

CH3

O

CH3

CH3

O

Si

H3C CH3

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH3CH2

CH2CH2

CH2CH2

N+

CH2

CH2CH2

OOR

OR

ORSi

H3C CH3

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH3CH2

CH2CH2

CH2CH2

N+

CH2

CH2CH2

SiOR

H3C CH3

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH2CH2

CH3CH2

CH2CH2

CH2CH2

N+

CH2

CH2CH2

OR

R = H, silicone rubber or Si CH2 CH2 CH2 N+

CH3

C18H37

CH3

OR

OR


93

photoelectron spectroscopy (XPS) using an S-Probe spectrometer (Surface Science Instruments,

Mountain View CA, USA). The elemental surface compositions were expressed in atomic %,

setting %C + %O + %N + %Si + %Cl to 100%. Zeta potentials of the surfaces were derived

from the pressure dependence of the streaming potentials employing a parallel plate flow

chamber [19]. The walls of the flow chamber were constituted by SR or QAS-coated SR sheets

fixed on Perspex plates (25 x 76 mm), separated by an 0.2 mm Teflon gasket, while two

rectangular platinum electrodes (5.0 x 25.0 mm) were located at both ends of a parallel plate

flow chamber [19]. Streaming potentials were measured during 1 h in PBS (pH 7.0), at ten

different pressures ranging from 37.5 to 150 Torr and each pressure was applied for 10 s in both

directions. Water contact angles were measured at room temperature with a home-made contour

monitor using the sessile drop technique. Advancing and receding angles were obtained by

keeping the needle in the water droplet after positioning on the surface and by carefully moving

the sample until the advancing angle was maximal. Each value was obtained by averaging results

of at least three droplets on one disc.

In vitro antimicrobial activity

Two SR and two QAS-coated SR discs were fixed in circular (diameter 8 mm) 0.5 mm deep

sections grooved in the middle of a 2 mm thick Perspex bottom plate of a parallel plate flow

chamber, which has previously been described in detail [20]. One SR and QAS-coated SR

disc were covered with pooled human plasma for 1 h. The chamber was filled with sterile

PBS and perfused for 30 min with PBS at a flow rate of 0.025 ml s-1. Subsequently, flow was

switched to a bacterial suspension of 3 x 108 cells ml-1 in PBS for 1 h. Then, flow was

switched for 30 min to PBS without bacteria to remove unbound organisms from the tubes

and the flow chamber at the same flow rate. The chamber was filled with 4 ml PBS

supplemented with 12 µl stock solution of live/dead baclight bacterial viability stain

(Molecular Probes, Oregon, USA), to determine the percentage of viable, adhering bacteria,

and from that moment on protected against light. After 30 min, 3 three-dimensional

fluorescent images (187.5 x 187.5 x 10 to 20 µm) were taken of each disc with a confocal

laser scanning microscope (CLSM, Leica TCS SP2, Heidelberg, Germany). Discs were

excited with 488 and 543 nm light and emitted light with wavelength 495 to 535 nm, due to

viable bacteria, was assigned the color green, while the light from 580 to 700 nm, arising from

non-viable organisms, was assigned the color red. Viability was defined as the percentage of

Chapter 7

94

viable, adhering bacteria relative to the total number of adhering bacteria. The experiments

were performed in duplicate.

In vivo antimicrobial activity

The SR and QAS-coated SR discs were washed in sterile PBS for 30 min. For preoperative

seeding, the discs were first incubated for 1 h in 5 x 109 CFU ml-1 S. aureus ATCC 12600 in

PBS, followed by 5 min in PBS without bacteria and finally 25 min in another batch of clean

PBS. Before implantation, the backs of the rats were shaved and disinfected with 0.5%

chlorhexidine in 70% ethanol, after induction of inhalation anesthesia with N2O/O2 (3/2) and

halothane. Four 1 cm incisions were made, two on either side of the middle line, at least 2 cm

apart. Subcutaneous pockets of at least 2 cm deep were created. A SR and QAS-coated SR

disc were inserted in the back pockets, while preoperatively seeded SR and QAS-coated SR

discs were inserted in the front pockets. The front pockets were closed with degradable suture

material. In the back pockets 0.2 ml 5 x 109 CFU ml-1 in PBS was injected over the implants,

and the wounds were closed. Four preoperatively seeded SR and QAS-coated SR discs were

not implanted, but added to 5 ml reduced transport fluid (RTF), for enumeration. Both at days

3 and 7, 4 rats were terminated with CO2. After skin disinfection and opening, the discs were

a-septically removed and added to 5 ml RTF.

For enumeration, discs in RTF were sonicated for 10 min. Then, the suspensions were

diluted and 100 µl of each dilution was streaked on blood agar to determine the number of

CFU. Discs were considered infected when adhering, viable bacteria were detected, regardless

of their number.

Results

The chemical and physical characteristics of SR and QAS-coated SR surfaces are summarized

in Table 1. The presence of QAS can be clearly seen from the increases in the %N and %Cl

relative to uncoated SR. Equilibrium water contact angles are hardly affected by the QAS-

coating, although the advancing contact angle increases upon QAS-coating whereas the

receding contact angle decreases. Most importantly, the zeta potential of SR, authentically

negative, becomes positive after coating. In CLSM, the QAS-coating shows as a slightly

green fluorescent layer, with occasionally about 5 micron deep cracks.


95

Table 1. Chemical surface composition (between brackets the theoretical composition based on molecular

composition), water contact angles (equilibrium, advancing and receding) and zeta potential in PBS of silicone

rubber (SR) and quaternary ammonium silanized silicone rubber (QAS-coated SR).

Surface property SR QAS-coated SR

%C 49 (50) 63 (80)

%O 26 (25) 19 (10)

%Si 25 (25) 14 (3.3)

%N 0 (0) 2.6 (3.3)

%Cl 0 (0) 2.3 (3.3)

Equilibrium water contact angle 111 ± 2 degrees 106 ± 3 degrees

Advancing water contact angle 117 ± 3 degrees 126 ± 4 degrees

Receding water contact angle 79 ± 4 degrees 57 ± 2 degrees

Zeta potential -15 mV +16 mV

The numbers of viable and non-viable adhering bacteria on SR and QAS-coated SR in vitro

after deposition for 1 h in the parallel plate flow chamber are presented in Figure 2, while the

resulting percentages viability are summarized in Table 2.

Table 2. The percentage viability (%) of adhering bacteria on silicone rubber (SR) and quaternary ammonium

silanized silicone rubber (QAS-coated SR) with and without adsorbed plasma proteins. Values are the average ±

SD of 6 images collected in 2 experiments, with separately cultured bacteria and differently prepared coatings.

Without adsorbed plasma proteins With adsorbed plasma proteinsBacterial strain

SR QAS-coated SR SR QAS-coated SR

S. aureus ATCC 12600 86 ± 3 0 85 ± 3 2 ± 3

S. epidermidis HBH2 102 94 ± 7 0.1 ± 0.2 90 ± 12 0.3 ± 0.4

E. coli O2K2 97 ± 1 37 ± 10 90 ± 5 26 ± 17

P. aeruginosa AK1 82 ± 5 15 ± 9 95 ± 2 19 ± 4

The great majority of organisms adhering on SR and SR after adsorption of plasma proteins is

viable (82% - 97%), whereas bacteria adhering on the QAS-coating are significantly less

viable (0% - 37%). Interestingly, adsorption of plasma proteins on the QAS-coating did not

shield the antimicrobial effect of the coating and viability remained low between 0.3% and

26%. The viability of bacteria adhering on the Perspex plate adjacent to the QAS-coated discs

was similar as of bacteria adhering elsewhere on the bottom plate of the flow chamber and on

SR discs, indicating that no antimicrobial QAS molecules were leaking from the QAS-

coating.

Chapter 7

96

S. aureus ATCC 12600

SR QAS SR + p QAS + p

Bac

teria

(106 c

m-2

)

0

1

2

3

4

5

6

7

8

S. epidermidis HBH2 102


E. coli O2K2


Bac

teria

(106 c

m-2

)

0

1

2

3

4

5

6

7

8

P. aeruginosa AK1


Figure 2. The numbers of adhering viable (black bars) and non-viable (white bars) bacteria on silicone rubber

(SR) and quaternary ammonium silanized silicone rubber (QAS) with and without adsorbed plasma proteins (SR

+ p and QAS + p). Error bars represent the SD over 6 images collected in 2 experiments, with separately cultured

bacteria and differently prepared coatings.


97

Table 3. Incidence of infections and numbers of CFU on silicone rubber (SR) and quaternary ammonium

silanized silicone rubber (QAS-coated SR) after subcutaneous implantation in rats, and after preoperative and

postoperative seeding of S. aureus ATCC 12600. Incidences are number of infected discs/number of implanted

discs and CFU are averages of the 4 samples ± SE (105 cm-2) explanted from 4 different rats.

Time Infection Preoperatively seeded Postoperatively seeded

SR QAS-coated SR SR QAS-coated SR

before Incidence 4/4 0/4 n.d. n.d.

CFU 48 ± 4 0 n.d. n.d.

3 days Incidence 4/4 1/4 4/4 4/4

CFU 6.3 ± 2.1 0.01 ± 0.01 3.1 ± 1.3 1.1 ± 1.0

7 days Incidence 3/4 0/4 3/4 2/4

CFU 3.0 ± 3.0 0 2.5 ± 1.6 0.6 ± 0.6n.d. = not determined

Table 3 summarizes the results of the in vivo evaluations of the QAS-coatings. Note that on

preoperatively seeded QAS-coated SR no detectable viable S. aureus were present before

implantation. Back pockets of all animals, containing SR discs, showed severe pus formation

at days 3 and 7, while the pockets with QAS-coated discs showed only mild pus formation.

Nearly all preoperatively seeded SR discs showed viable staphylococci after 3 and 7 days,

which is completely opposite to QAS-coated discs, showing almost no viable organisms. All

postoperatively seeded SR and QAS-coated discs showed viable organisms after 3 days,

although fewer viable bacteria were harvested from the QAS-coated discs than from the SR

discs. After 7 days the SR discs (3 out of the 4) showed similar numbers of CFU than after 3

days. Two QAS-coated discs showed viable bacteria, albeit in lower numbers than after 3

days.

Discussion

In this study silicone rubber was coated with 3−(trimethoxysilyl)propyldodecyldimethyl

ammonium chloride (QAS), and the antimicrobial activity of this coating was evaluated, both

in vitro and in vivo. The QAS-coating showed antimicrobial activity towards all tested

bacterial strains in vitro, which was confirmed in vivo for one selected S. aureus strain.

QAS molecules in solution can interact with the lipid bilayer structures of microbial

cell membranes [21]. From our study, we conclude that immobilized QAS molecules still

Chapter 7

98

interact with the cell membranes of adhering bacteria, presumably causing membrane leakage

and cell death.

Isquith et al. were the first to report antimicrobial activity of QAS-coated glass and

cotton towards Streptococcus faecalis, E. coli, fungal spores and algi [22], although their

isolation method was questionable. Others reported that Gram-negative bacilli were not

affected by QAS [16]. Our study clearly demonstrates that also the viability of Gram-negative

bacilli is affected by QAS, but that higher numbers of negatively charged Gram-negative

bacteria adhere to the positively charged QAS-coating than to the bare SR, which may easily

lead to a misinterpretation of the effect of the coating on viability.

Flemming et al. [23] also found good antimicrobial activity toward S. aureus of

polyurethanes functionalized with methyl and ethyl quaternary ammonium iodide groups.

However, the iodide counter ion played an important role as the same polymer with chloride

counter ions displayed less activity. Furthermore, agar inhibition zones were seen for

polyurethanes functionalized with iodide, indicating that antimicrobial agents were leaking

from the polymers. Albumin treatment of polyurethanes with ethyl quaternary ammonium

chloride led to a loss of antimicrobial activity [23], which is opposite to this study as

preadsorption of human plasma proteins did not significantly reduce the antimicrobial activity

of the QAS-coating. Likely, the long aliphatic chains of adsorbed QAS molecules sterically

hinder association of proteins with the quaternary ammonium groups, leaving enough

quaternary ammonium groups available for interaction with the cell membranes of adhering

bacteria.

As a most relevant feature of this study, both in vitro and in vivo antimicrobial activity

of QAS-coated silicone rubber is demonstrated. Therewith, QAS-coatings may significantly

contribute to prevent biomaterial-centered infections, at least by S. aureus. The application of

positively charged biomaterial surfaces to prevent infection is a 180 degrees turn in thinking,

as current research has been directed toward designing non-adhesive surfaces. Positively

charged surfaces are strongly adhesive toward negatively charged bacteria, but evidently the

positive charge inhibited biofilm formation to proceed from the stage of initial adhesion

toward growth.


99

Conclusion

3-(trimethoxysilyl)propyldimethyloctadecylammonium chloride coating of silicone rubber

yields a positively charged, highly adhesive surface with antimicrobial effects towards

adhering Gram-positive and Gram-negative bacteria. The antimicrobial effects also exist

under in vivo conditions, at least against S. aureus in the rat. QAS-coated biomaterial implants

thus seem promising in preventing biomaterial-centered infection in man.

References


Science 237, 1588-95.












6. Hugo, W.B. and Russel, A.D. (1992). Types of antimicrobial agents. In Principles and practice of

disinfection, preservation and sterilization. Edited by Russel, A.D., Hugo, W.B. and Ayliffe, G.A.J.

(Blackwell Scientific Publications, Oxford) p. 7-68.

7. Gould, I.M. (1999). A review of the role of antibiotic policies in the control of antibiotic resistance.

Journal of Antimicrobial Chemotherapy 43, 459-65.



9. Gottenbos, B., Grijpma, D.W., Van der Mei, H.C., Feijen, J. and Busscher, H.J. (2001). Antimicrobial

effects of positively charged poly(methacrylates) on adhering Gram-negative and Gram-positive bacteria.

Journal of Antimicrobial Chemotherapy (submitted).




Chapter 7

100

11. Ahlstrom, B., Chelminska-Bertilsson, M., Thompson, R.A. and Edebo, L. (1995). Long-chain

alkanoylcholines, a new category of soft antimicrobial agents that are enzymatically degradable.

Antimicrobial Agents and Chemotherapy 39, 50-5.

12. Ahlstrom, B., Thompson, R.A. and Edebo, L. (1999). The effect of hydrocarbon chain length, pH, and

temperature on the binding and bactericidal effect of amphiphilic betaine esters on Salmonella

typhimurium. APMIS 107, 318-24.

13. Calas, M., Cordina, G., Bompart, J., Ben-Bari, M., Jei, T., Ancelin, M.L. and Vial, H. (1997).

Antimalarial activity of molecules interfering with Plasmodium falciparum phospholipid metabolism.

Structure-activity relationship analysis. Journal of Medical Chemistry 40, 3557-66.

14. Lindstedt, M., Allenmark, S., Thompson, R.A. and Edebo, L. (1990). Antimicrobial activity of betaine

esters, quaternary ammonium amphiphiles which spontaneously hydrolyze into nontoxic components.

Antimicrobial Agents and Chemotherapy 34, 1949-54.

15. Zhisheng Chen, C., Beck Tan, N.C. and Cooper, S.L. (1999). Incorporation of

dimethyldodecylammonium chloride functionalities onto poly(propylene imine) dendrimers significantly

enhances their antibacterial properties. Chemical Communications, 1585-6.

16. Murray, P.R., Niles, A.C. and Heeren, R.L. (1988). Microbial inhibition on hospital garments treated

with Dow Corning 5700 antimicrobial agent. Journal of Clinical Microbiology 26, 1884-6.




18. Silver, J.H., Lin, J.C., Lim, F., Tegoulia, V.A., Chaudhury, M.K. and Cooper, S.L. (1999). Surface

properties and hemocompatibility of alkyl-siloxane monolayers supported on silicone rubber: effect of alkyl

chain length and ionic functionality. Biomaterials 20, 1533-43.





21. Merianos, J.J. (1991). Quaternary ammonium antimicrobial compounds. In Disinfection, sterilization, and

preservation . Edited by Block, S.S. (Lea & Febiger, Philadelphia) p. 225-55.

22. Isquith, A.J., Abbott, E.A. and Walters, P.A. (1972). Surface-bonded antimicrobial activity of an

organosilicon quaternary ammonium chloride. Applied Microbiology 24, 859-63.

23. Flemming, R.G., Capelli, C.C., Cooper, S.L. and Proctor, R.A. (2000). Bacterial colonization of

functionalized polyurethanes. Biomaterials 21, 273-81.

101

General discussion

In this thesis antimicrobial biomaterial surfaces and methods to study the antimicrobial

activity of these surfaces were developed. Antimicrobial biomaterial surfaces were defined as

surfaces on which the surface growth of adherent bacteria is inhibited without the use of

antimicrobial agents leaking from the surface. Inhibition of surface growth should be achieved

by changing the physico-chemical properties of the biomaterial surface, or by binding

antimicrobial agents covalently to the biomaterial surface.

Changing surface properties and covalent binding of antimicrobial agents

One approach to change the surface properties of biomaterials is to synthesize polymers with

the desired surface properties, and subsequently coat the biomaterial surface with this

polymer. Adhesive Van der Waals forces and, ideally, adhesive electrostatic forces should

keep the polymer chains attached to the surface. The positively charged poly(methacrylate),

synthesized as described in Chapters 4 and 5, can thus be coated on most negatively charged

surfaces, which are virtually all commonly used biomaterial surfaces. Also, biomaterial

implants with complex geometry can be easily coated with these polymers by dip coating [1].

Poly(methacrylates) can be synthesized with a large variety of functional groups, as many

different methacrylate monomers are readily available. However, the percentage of ionic or

hydrophilic functional groups in the polymers must not be too high, as these increase the

solubility of the polymer in water, which results in a lower durability of the coatings. In

addition the mechanical properties of the polymer coating must be suitable for the application.

Our positively charged poly(methacrylate) has a glass transition temperature of 60°C,

meaning that at room temperature the coating might be too stiff to be coated e.g. on silicone

rubber catheters.

8

Chapter 8

102

A second approach was to bind antimicrobial functional groups covalently to the

biomaterial surface with silane coupling agents. Also silane coupling reagents are readily

available with a large variety of functional groups. Silane coupling reagents react with

hydroxyl groups, which are present on e.g. glass and cotton, to form a silicium ether bond [2].

Hydroxyl groups are also introduced on silicone rubber by oxidizing the surface with argon

plasma etching [3]. Silanes can also react with itself, forming a crosslinked polymer layer on

the biomaterial surface. In in vitro experiments as described in Chapter 7, this layer proved to

be durable for at least some hours. The durability for prolonged implantation remains to be

tested. In vivo sterile pus formation was found around the coated silicone rubber, indicating

fragmentation of the coating, which can induce a strong foreign body reaction. In the warm,

wet, oxygenated in vivo environment silicium ether bonds might not be stable towards

hydrolysis.

In a pilot study that was not reported in this thesis, another approach to introduce

covalently bonded functional groups to biomaterials was tested, so-called graft co-

polymerization. In this technique, first peroxide groups are introduced on the biomaterial

surface by gas plasma treatment, followed by exposure to oxygen. Subsequently, these

peroxide groups are used to initiate polymerization of acrylate monomers, resulting in

poly(acrylate) chains covalently bound to the biomaterial surface. We succeeded in graft co-

polymerizing a small amount of polyacrylamide onto polyethylene. However, as optimization

of the process is very time consuming due to the many different parameters involved, it was

chosen to continue first with the methods described above, as faster results could be expected,

although eventually graft co-polymerization can result in more durable coatings.

In Table 1 the respective advantages and disadvantages of the three described methods

are summarized.

Table 1. Characteristics of the various approaches to introduce functional groups on biomaterial surfaces. + =

good, +/- = intermediate and - = bad.

Property Polymer coating Silanization Graft co-polymerization

Durability +/- +/- +

Applicable biomaterial range + - +/-

Difficult implant geometry + +/- +/-

Functionality range + + +

Mechanical properties +/- + +

General discussion

103

Methods to study the antimicrobial effects of biomaterial surfaces in vitro

In this thesis we used three different in vitro methods to study the antimicrobial activities of

our antimicrobial biomaterial surfaces. Table 2 shows the various parameters that can be

determined, and some characteristics of the respective methods. The parallel plate flow

chamber system described in Chapter 2 can be used to study bacterial surface growth in situ

on translucent or reflecting biomaterials. This method is suitable for evaluating antimicrobial

surfaces that decrease surface growth, but can also determine the killing efficacy of surfaces,

covalently linked with antimicrobial agents, although non-growing adherent cells are not by

definition non-viable. They can also be incapable of growing on the surface for another

reason.

Table 2. Comparison of three in vitro methods to study the antimicrobial activity of biomaterial surfaces, i.e.

surface growth measurements of adherent bacteria in the parallel plate flow chamber (PPFC), viability staining

of adherent bacteria in the PPFC and colony forming units (CFU) determination. Shown are the various

parameters that can be determined, the applicability and the reliability.

Parameter Growth PPFC Viability stain PPFC CFU determination

Initial adhesion rates Yes Yes No

Surface growth rates Yes No No

Desorption rates Yes No No

Viable bacteria Yes Yes Yes

Non-viable bacteria Yes/No* Yes No

Viable but non-growing bacteria Yes/No* No No

Possible with all materials No Yes Yes

Clear indisputable method Yes Yes No* No distinction can be made between these two states of adherent bacteria.

A faster way to measure the viability of adhering bacteria in the parallel plate flow chamber is

to add a viability stain in stead of growth medium after the initial adhesion phase. The

live/dead baclight bacterial viability stain (Chapter 7) can be used, where viable bacteria

appeared green fluorescent, whereas dead bacteria with disrupted cell membranes colored red

fluorescent. Also staining with 5-cyano-2,3-ditolyl tetrazolium chloride (CTC) is possible, as

used by Habash et al. [4]. Here, only the actively respiring bacteria are stained red

fluorescent. We also used this staining in a few experiments (not reported) and found for S.

epidermidis HBH2 102 on silicone rubber that after 1 h of initial adhesion in buffer only few

Chapter 8

104

adhering bacteria were actively respiring, usually those that were present in groups. This

corresponded to what we found in Chapter 4, where only 7% of S. epidermidis HBH2 102

cells on PMMA/MAA was growing. However the CTC stains did not correspond at all with

the live/dead viability stains in Chapter 7 where this bacterial strain showed a viability of 94%

on silicone rubber after 1 h of initial adhesion. This indicates that S. epidermidis with intact

cell membranes are not necessarily actively respiring or capable of surface growth within the

first hours after administration of growth medium. These bacteria might have been in the so-

called viable but non-culturable state. However, for P. aeruginosa AK1 the percentage of

actively respiring bacteria on silicone rubber as determined with CTC was found to be much

higher, corresponding to the high percentage of growing cells on PMMA/MAA (Chapter 4)

and the cells with intact cell membranes on silicone rubber (Chapter 7).

The third method is to let the bacteria adhere to the biomaterial surface and

subsequently detach them from the surface by ultrasonic treatment of the biomaterial and

determine the number of detached colony forming units (CFU). This method is extensively

used in the literature, but only reveals the number of viable detachable adhering bacteria (see

Table 2). Another disadvantage is that some operations are debatable. The biomaterial passes

an air/liquid interface, which is known to exert a high shear force on the bacteria, which could

push the bacteria from the surface before the ultrasonic treatment. Also, not all the adhering

bacteria may be removed, especially on extremely adhesive positively charged surfaces.

In conclusion, determination of surface growth combined with a stain to test the

viability of non-growing cells in a parallel plate flow chamber is the most suitable method to

evaluate antimicrobial biomaterial surfaces.

Animal models to study biomaterial-centered infections

Direct/indirect immediate local infections

In immediate local infection models the bacteria are implanted together with the biomaterial.

A direct infection model was applied in Chapters 5 and 7, where the bacteria were seeded

directly on the biomaterial surface before subcutaneous implantation in the rats. In this way

no conditioning film was present between bacteria and biomaterial, as was the case in the in

vitro experiments. This model resembles in the clinic the cases where bacteria are seeded

before or during insertion of the implant from the air, the surgeon or the patient’s skin. Our

results showed that only P. aeruginosa AK1 was capable of growing on the implant surface.

General discussion

105

On silicone rubber the numbers of S. aureus ATCC 12600 had decreased 8 fold in 3 days and

16 fold in 1 week, while 1 out of 4 implants was sterile after 1 week. The numbers of E. coli

O2K2 decreased fast, a 100 fold in 2 days. The same was found for S. epidermidis HBH2 102,

where, after seeding of 106 CFU cm−2 on silicone rubber, 6 out of 7 implants (in 7 rats) were

sterile after 3 days, while only 103 CFU cm−2 were harvested from the infected implant (non-

reported experiments).

The observed differences between P. aeruginosa and the other bacterial strains were

probably due to P. aeruginosa’s capability to form slime and fast biofilm formation. In vitro

(Chapters 3, 4) P. aeruginosa AK1 formed a monolayer biofilm within 7 h, while this took

longer for the staphylococci and especially for E. coli O2K2 because of its high desorption

rates from the surface. S. epidermidis HBH2 102 is also a good slime producer, but this is

much more dependent on the nutrient conditions, as was observed in Chapter 3. Besides slime

formation and fast surface growth, other virulence factors of P. aeruginosa AK1, like

endotoxin formation and motility can be important. Advantages for causing infection of S.

aureus over S. epidermidis are faster surface growth rates (Chapter 4), production of

exotoxins and circumventing phagocytosis by elimination of antibodies and phagocytes.

An indirect local infection model was also applied in Chapter 7. Here, 109 CFU S.

aureus were added to the implant site, after inserting the implants, to study the effect of a

conditioning film of plasma proteins on the antimicrobial biomaterial surface before adhesion

of the bacteria. This animal model is most often chosen in literature to study biomaterial-

centered infections. It resembles the clinical cases where the bacteria are seeded from the

operation wound after the insertion of the implant, which is generally believed to be the most

common cause of biomaterial-centered infections. In our study all implant sites showed

clinical signs of infection, and bacteria were recovered from 7 out of 8 silicone rubber

implants, which was equal to the direct infection model. Thus, from these results, the direct

and indirect model have a similar likelihood of representing the pathogenesis of biomaterial-

centered infections.

In non-reported experiments the addition of 2 x 108 CFU S. aureus led to sterile

implants after 1 week (two rats). This corresponds to the literature, where addition of 109 CFU

S. aureus to polyethylene implants in rats resulted in infection of most of the implants at 7

days, while addition of 108 CFU showed no evidence of infection [5]. However, it was shown

that seeding biomaterials with as low as 100 to 1000 CFU S. aureus could produce clinical

biomaterial-centered infections in mice and rabbits [6, 7]. Also Zimmerli et al. [8] found that

Chapter 8

106

an avirulent strain, S. aureus Wood 46, could produce a biomaterial-centered infection with

only 100 CFU in >95% of their so-called tissue cages implanted subcutaneously in guinea

pigs. This corresponds to an experiment in man where 100 CFU of S. aureus were enough to

cause infections in the presence of a suture [9]. These data indicate that rats might react

differently to the induced biomaterial-centered infections than other animals, including man.

While it was shown in man and guinea pigs that the surroundings of biomaterials are in fact

immuno-compromized, rats may have other immunologic pathways to eradicate bacteria from

inert surfaces. Also pointing to this direction are the reported major differences between the

foreign body reaction of rats and mice [10].

Late hematogenous infections

An advantage of antimicrobial biomaterial surfaces over antibiotic releasing materials should

be the prolonged duration of the protection against infections. The use of such biomaterial

surfaces could protect implants against postulated late hematogenous seeding of

microorganisms, if at all occurring. In Chapter 6 was studied if subcutaneously implanted

biomaterials in rats could be infected by hematogenous seeding after 4 weeks of implantation.

It was found that none of the implants were infected by intravenously injected bacteria or

translocated intestinal bacteria. These results can be explained by the fact that the bacterial

concentration near the subcutaneous implants was much lower than in the local infection

models, due to dilution effects, filtering of the blood by organs (injected staphylococci did

infect all kidneys) and tissues and the action of the immune system. Moreover, at this late

stage the subcutaneous biomaterials were covered with host cells, therewith reducing the

chances of bacterial adhesion.

Surprisingly, in these experiments, 5% of the implants were infected, possibly due to

perioperative contamination, as the bacterial strains did not correspond with the injected or

translocated strains. It might be that these low virulent (no clinical infection signs) bacteria,

probably originating from the commensal skin flora of the rats, are immunotolerated by the

rats, thus being able to colonize the biomaterial surface without being eradicated, a hypothesis

proposed in Chapter 1.

From these results can be concluded that late hematogenous infections are much less

likely to occur than immediate local infections. Also in the literature little evidence can be

found for the occurrence of late hematogenous infections. This indicates that most

General discussion

107

biomaterial-centered infections can be prevented with antimicrobial biomaterial surfaces that

are active for only a short period after implantation.

Mechanisms of antimicrobial activity of biomaterial surfaces

As with all antimicrobial agents, the various types of bacteria studied in this thesis reacted

differently towards the respective antimicrobial and non-antimicrobial biomaterial surfaces. It

was observed that growth of P. aeruginosa AK1, a Gram-negative rod, was hindered on

surfaces onto which it attached strongly, resulting in slower surface growth rates (Chapter 3)

(see Figure 1).

BiomaterialBiomaterial

A B

Figure 1. Schematic presentation of surface growth of rod shaped P. aeruginosa and E. coli (A) and round

shaped S. aureus and S. epidermidis (B). Note that the rod shaped bacteria have to overcome the physical

attraction to the biomaterial surface in order to elongate and divide.

Ultimately on positively charged poly(methacrylate) the attraction to the surface was so

strong, that none of the adhering P. aeruginosa AK1 could grow (Chapter 4). It became clear

in Chapter 5 that attraction to the surface was not the only growth inhibition mechanism.

When the adhering bacteria were detached from the positively charged surface only half of

the bacteria were able to form a colony on agar, i.e. were viable. The positive charge in the

polymers was induced by quaternary ammonium groups, which can be toxic for bacteria and

probably killed half of the adhering bacteria. Also the quaternary ammonium containing

silane coating was shown to be toxic for 75% of the adhering P. aeruginosa AK1 cells

(Chapter 7). Like on the positively and negatively charged poly(methacrylates) (Chapter 5),

the number of viable bacteria were the same on the positively charged silane coated silicone

rubber and the negatively charged plain silicone rubber, as initial adhesion was much higher

to positively charged surfaces. In a non-reported pilot experiment, surface growth of adhering

P. aeruginosa AK1 on the same quaternary ammonium silane coated glass was measured in a

parallel plate flow chamber, exactly as it was done on the positively charged

Bacterium

Chapter 8

108

poly(methacrylates) (Chapter 4). We saw that none of the adhering bacteria were able to grow

on this surface, indicating that the surface growth of adhering viable bacteria was probably

also inhibited by the strong attraction of a positively charged surface. However, this physical

surface growth inhibition can be overcome by P. aeruginosa AK1 in vivo, as was shown in

Chapter 5.

The antimicrobial effects of charged surfaces were also tested for another Gram-

negative rod shaped bacterium: E. coli O2K2 (Chapters 4,5,7). In vitro the same effects were

seen as those described above for P. aeruginosa AK1, i.e. no surface growth, and killing of

adhering bacteria on the positively charged surface. The differences in initial adhesion

between positively and negatively charged surfaces were larger for E. coli, as this strain has a

higher negative surface charge. In vivo, however, the physical inhibition of surface growth

appeared to be effective (Chapter 5). As E. coli O2K2 is not a very good slime and biofilm

producer, it was proposed that viable adhering P. aeruginosa AK1 cells could overcome the

physical surface growth inhibition by forming slime between cells and the biomaterial

surface, therewith loosening the bond between them, in contrast to E. coli.

The staphylococci studied in this thesis reacted in a different pattern towards

biomaterial surfaces. Surface growth rates of S. epidermidis HBH2 102 correlated with the

surface hydrophobicity, although this effect was only seen under low nutrient conditions, and

polypropylene was an exception to this rule (Chapter 3). Interestingly, corresponding to the

very slow in vitro surface growth on glass and polypropylene, in Chapter 6 only glass and

polypropylene showed no perioperatively introduced biomaterial-centered infections,

although, due to the low number of infections found, this relation is not statistically

significant.

The staphylococci were probably affected by the positively charged

poly(methacrylate), as the percentage of growing bacteria was half of that on normal PMMA

(Chapter 4). This corresponds with the Gram-negative rod shaped bacteria, that were

recovered from the positively charged surface, from which the viability was half of those

recovered from PMMA (Chapter 5), indicating that the quaternary ammonium groups were

also toxic for the staphylococci. However, surface growth of staphylococci was not inhibited

by strong physical attraction, as these round bacteria do not need to elongate much before cell

division (see Figure 1). In Chapter 7 almost all adhering staphylococci were killed by the

positively charged silane coated silicone rubber, even when the surface was precoated with

plasma proteins. In vivo, this activity was retained, although here probably plasma proteins

General discussion

109

did cover a considerable part of the antimicrobial groups, resulting in a lower killing efficacy.

The quaternary ammonium groups in this coating are probably more active, because the long

hydrophobic alkyl chain can interfere with the lipids of the cell membrane, thereby disrupting

its integrity. The Gram-negative bacteria might be less affected by this compound as these

bacteria have double cell membranes.

Future research

It would be interesting to test the hypothesis that perioperatively introduced biomaterial-

centered infections in rats are caused by immunotolerated microorganisms. To this end the

percentage of perioperative infections could be increased by rubbing the implants over the

skin of the rats before implantation. Immunotolerance of the rats for the infecting

microorganisms could be determined by the method of Duchmann et al. [11].

Positively charged coatings might also be effective in the prevention of urinary tract

catheter infections, as E. coli is often involved in these infections. The adhesive character of

the surface might also inhibit the migration of bacteria along the catheter surface from the

outside of the body to the bladder, were bacteriuria is caused. Also the quick migration, so-

called swarming, of proteus spp., the second most common pathogen in these infections,

might not be possible on these coatings. To test this, adhesion, growth and motility could be

studied in urine in the parallel plate flow chamber. In this application the coatings should also

be durable for the insertion period, which should also be tested. An approach to make more

durable coatings could be the graft co-polymerization of methacrylates functionalized with

quaternary ammonium groups.

Another interesting approach to make specific adhesive surfaces for growth inhibition

of rod-shaped bacteria, might be coating of biomaterial surfaces with immunoglobulins. When

monoclonal immunoglobulin G (IgG) directed against for example P. aeruginosa is bound

with the Fc side to the biomaterial, in theory this surface should be very adhesive for the

bacteria, thus preventing their proliferation on the surface. An indication that this might work

can be obtained from the study of Poelstra et al. [12] where adsorbed pooled human IgG on

the substratum surface decreased the surface growth rates of P. aeruginosa by 15%.

Studying the relationship between adhesion and growth can be also very interesting on

metals, as their surface charge can be easily varied using an electrical power source [13]. In

Chapter 8

110

this way it might be possible to determine which attraction force is needed to slow down

growth rates of for example P. aeruginosa.

References



2. Plueddemann, E. P. (1982). Silane Coupling Agents, Plenum Press, New York.







5. Sclafani, A.P., Thomas, J.R., Cox, A.J. and Cooper, M.H. (1997). Clinical and histologic response of

subcutaneous expanded polytetrafluoroethylene (Gore-Tex) and porous high-density polyethylene

(Medpor) implants to acute and early infection. Archives of Otolaryngology: Head and Neck Surgery 123,

328-36.

6. Merritt, K., Hitchins, V.M. and Neale, A.R. (1999). Tissue colonization from implantable biomaterials

with low numbers of bacteria. Journal of Biomedical Materials Research 44, 261-5.

7. Poelstra, K.A., Barekzi, N.A., Grainger, D.W., Gristina, A.G. and Schuler, T.C. (2000). A novel spinal

implant infection model in rabbits. Spine 25, 406-10.

8. Zimmerli, W., Waldvogel, F.A., Vaudaux, P. and Nydegger, U.E. (1982). Pathogenesis of foreign body

infection: description and characteristics of an animal model. Journal of Infectious Diseases 146, 487-97.

9. Elek, S.D. and Conen, P.E. (1957). The virulence of Staphylococcus pyogenes for man. A study of the

problems of wound infection. British Journal of Experimental Pathology 38, 573-86.

10. Khouw, I.M.S.L., Van Wachem, P.B., Molema, G., Plantinga, J.A., De Leij, L.F.M.H. and Van Luyn,

M.J.A. (2000). The foreign body reaction to a biodegradable biomaterial differs between rats and mice.

Journal of Biomedical Materials Reseach 52, 439-46.




12. Poelstra, K.A., Van der Mei, H.C., Gottenbos, B., Grainger, D.W., Van Horn, J.R. and Busscher, H.J.

(2000). Pooled human immunoglobulins reduce adhesion of Pseudomonas aeruginosa in a parallel plate

flow chamber. Journal of Biomedical Materials Research 50, 224-32.

13. Poortinga, A.T., Bos, R. and Busscher, H.J. (2000). Controlled electrophoretic deposition of bacteria to

surfaces for the design of biofilms. Biotechnology and Bioengineering 67, 117-20.

111

Summary

One of the major drawbacks in the use of biomedical materials is the occurrence of

biomaterial-centered infections. After implantation, the host interacts with a biomaterial by

forming a conditioning film on its surface and an immune reaction towards the foreign

material. When microorganisms reach a biomaterial surface they can adhere to it. Adhesion of

microorganisms to an implant is mediated by their physico-chemical surface properties and

the properties of the biomaterial surface itself. Subsequent surface growth of the

microorganisms will lead to a mature biofilm and infection, which is difficult to eradicate by

antibiotics. Chapter 1 gives an overview of the mechanisms involved in biomaterial-centered

infection and the possible strategies to prevent these infections.

Chapter 2 describes the parallel plate flow chamber system. With this system the

different steps in biofilm formation, i.e. conditioning film formation, initial bacterial

adhesion, bacterial surface growth and bacterial detachment can be modeled and monitored in

situ. Examples are given of studies concerning the influence of a plasma conditioning film on

initial adhesion, the influence of biomaterial surface properties on surface growth and the

influence of surface active substances on detachment of biofilm bacteria.

The infection risk of biomaterial implants varies between different materials and is

determined by an interplay of adhesion and surface growth of the infecting organisms. In

Chapter 3, we compared initial adhesion and surface growth of Staphylococcus epidermidis

HBH2 102 and Pseudomonas aeruginosa AK1 on poly(dimethylsiloxane), Teflon, polyethylene,

polypropylene, polyurethane, poly(ethylene terephthalate), poly(methyl methacrylate) and glass.

Initial adhesion was measured in situ in a parallel plate flow chamber with microorganisms

suspended in phosphate buffered saline, while subsequent surface-growth was followed in full

and in 20 times diluted growth medium. Initial adhesion of both bacterial strains was similar to

all biomaterials. In full growth medium, generation times of surface growing S. epidermidis

ranged from 17 to 38 min with no relation to wettability, while in diluted growth medium

generation times increased from 44 to 98 min with increasing surface wettability. For P.

aeruginosa no influence of surface wettability on generation times was observed, but

generation times increased with decreasing desorption rates, maximal generation times being

47 min and minimal values down to 30 min. Generally generation times of adhering bacteria

Summary

112

were shorter than of planktonic bacteria. In conclusion, surface-growth of initially adhering

bacteria is influenced by biomaterials surface properties to a greater extent than initial adhesion.

In chapter 4, the antimicrobial effects on adhering bacteria of a positively charged

poly(methacrylate) surface (zeta potential +12 mV) are compared with those of negatively

charged poly(methyl methacrylate) (-12 mV) and a highly negatively charged poly(methacrylate)

(-18 mV) surface. Initial adhesion of Staphylococcus aureus ATCC 12600, S. epidermidis HBH2

102, Escherichia coli O2K2 and P. aeruginosa AK1 to these biomaterial surfaces was measured

in a parallel plate flow chamber in phosphate buffered saline, while subsequently adhering

bacteria were allowed to grow by perfusing the flow chamber with growth medium. On the

positively charged surface adhesion was fastest, but subsequent surface growth, however, was

absent for the Gram-negative strains. On the negatively charged surfaces, despite a slower initial

adhesion, surface growth of the adhering bacteria was exponential for both Gram-positive and

Gram-negative strains. These results suggest that positively charged biomaterial surfaces exert an

antimicrobial effect on adhering Gram-negative bacteria, but not on Gram-positive ones.

The infection rate of differently charged poly(methacrylates) in rats was determined in

Chapter 5. To this end, 2 x 106 cm-2 E. coli O2K2 or 2 x 104 cm-2 P. aeruginosa AK1 were

seeded on glass discs with three differently charged poly(methacrylates) coatings in a parallel

plate flow chamber. Three rats received six subcutaneous discs (two discs of each charge variant)

seeded with E. coli, while three other rats received discs seeded with P. aeruginosa. The

numbers of viable bacteria on the surfaces were determined 48 h after implantation. Viable E.

coli were absent on 50% of all positively charged discs, while the negatively charged discs were

all colonized by E. coli. P. aeruginosa, however, were isolated from both positively and

negatively charged discs. Probably, P. aeruginosa can circumvent the antimicrobial effect of the

positive charge through the formation of an extracellular polymer layer.

Chapter 6 describes a study of the pathogenesis of late biomaterial-centered infection.

Subcutaneously implanted biomaterials in rats were hematogenously challenged with bacteria

4 weeks after implantation. Bacteria were spread by either intravenous injection or stimulation

of bacterial translocation. It was found that none of the biomaterials was infected by

hematogenous spread, whereas 5% of the implants were infected by perioperative

contamination. We conclude that late hematogenous infection of subcutaneous biomaterials

does not occur in the rat. Also in man, there are growing doubts whether implants actually

become infected through hematogenous routes or whether late infections are caused by

delayed appearance of perioperatively introduced bacteria.

Summary

113

In Chapter 7 the antimicrobial activity of silicone rubber with a covalently coupled 3-

(trimethoxysilyl)-propyldimethyloctadecylammonium chloride (QAS) coating was studied in

vitro and in vivo. Gram-positive S. aureus ATCC 12600, S. epidermidis HBH2 102, and

Gram-negative E. coli O2K2 and P. aeruginosa AK1 were seeded on the silicone rubber with

and without QAS-coating, in the absence or presence adsorbed human plasma proteins. The

viability of the adherent bacteria was determined using a live/dead fluorescent stain and a

confocal laser scanning microscope. The coating reduced the viability of adherent

staphylococci from 90% to 0%, and of Gram-negative bacteria from 90% to 25%, while the

presence of adsorbed plasma proteins had little influence. The biomaterials were also

subcutaneously implanted in rats for 3 or 7 days, while preoperatively or postoperatively

seeded with S. aureus ATCC 12600. Preoperative seeding resulted in infection of 7 out of 8

silicone rubber implants against 1 out of 8 QAS-coated silicone rubber implants.

Postoperative seeding resulted in similar infection incidences on both implant types, but the

numbers of adhering bacteria were 70% lower on QAS-coated silicone rubber. It was

concluded that QAS-coated silicone rubber shows antimicrobial properties against adhering

bacteria, both in vitro and in vivo.

Finally, in Chapter 8 the methods used to introduce functional groups and to evaluate

antimicrobial biomaterial surfaces in vitro are compared. The differences between different

bacterial strains in ability to produce biomaterial-centered infections in vivo in the direct local

infection model are discussed, and suggestions are made for further research.

115

Samenvatting voor niet ingewijden

Door de toenemende vergrijzing van de bevolking stijgt het gebruik van lichaamsvreemde

materialen voor herstel en ondersteuning van lichaamsfuncties, de zogenaamde biomaterialen.

Een veelvoorkomend probleem bij het gebruik van biomedische implantaten en hulpmiddelen,

zoals bijvoorbeeld heup protheses, kunst hartkleppen of urineweg katheters, is infectie. Deze

infecties ontstaan omdat eencellige organismen, meestal bacteriën (deze hebben een grootte

van ongeveer één duizendste millimeter), hechten aan het oppervlak van deze

lichaamsvreemde materialen. Het afweersysteem van de patiënt kan deze gehechte bacteriën

moeilijk uitschakelen, omdat de bacteriën aan de ene kant beschermd worden door het

biomateriaal oppervlak, en aan de andere kant door een slijmlaag, die ze zelf aanmaken. De

gehechte bacteriën kunnen zich hierdoor gemakkelijk ongehinderd vermeerderen, wat ze doen

door zich steeds opnieuw in tweeën te delen op het biomateriaal oppervlak, een proces wat we

oppervlakte groei genoemd hebben. Zo onstaat een laag bacteriën, een zogenaamde biofilm,

op het implantaat, en in dit stadium spreken we van een biomateriaal-gerelateerde infectie.

Door deze infectie kan het implantaat slecht gaan functioneren, terwijl de patiënt pijn krijgt,

ziek wordt en uiteindelijk zelfs kan overlijden. Omdat bacteriën in een biofilm ongevoelig

zijn voor antibiotica, is vaak de enige remedie tegen biomateriaal-gerelateerde infecties het

verwijderen van het implantaat of hulpmiddel (Hoofdstuk 1). Het is dus beter om dit soort

infecties te voorkomen, door biomaterialen te gebruiken waarop bacteriën geen biofilm

kunnen vormen. Het doel van het onderzoek beschreven in dit proefschrift was te

onderzoeken hoe we het oppervlak van biomaterialen moeten veranderen om de groei van

gehechte bacteriën tegen te gaan.

Hiervoor werd eerst een methode bedacht om de hechting en groei van gehechte

bacteriën te meten (Hoofdstuk 2). De hechting van bacteriën werd gemeten door met een

microscoop op het biomateriaal oppervlak te kijken en bacteriën in een vloeistof zonder

voedingsstoffen langs te laten stromen. Vervolgens werd de oppervlakte groei van de

gehechte bacteriën gemeten door vloeibaar voedsel langs te laten stromen. Ook het eventueel

weer loslaten van bacteriën van het biomateriaal oppervlak kon op deze wijze bekeken

worden.

Met deze methode werd de hechting en groei gemeten van twee verschillende

bacteriesoorten op acht verschillende, veel gebruikte biomaterialen (Hoofdstuk 3). Hieruit

Samenvatting

116

bleek dat het aantal bacteriën wat hechtte, weinig verschilde tussen de biomaterialen, terwijl

de oppervlakte groei juist wel veel verschilde. De onderzochte bolvormige bacterie

(Staphylococcus epidermidis) groeide sneller op meer water afstotende (hydrofobe)

biomaterialen, waarschijnlijk door een betere beschikbaarheid van voedsel op deze

materialen. Dit gold echter niet voor de onderzochte staafvormige bacterie (Pseudomonas

aeruginosa). Deze groeide langzamer op biomaterialen waarvan er minder bacteriën van het

oppervlak loslieten, m.a.w. biomaterialen waar de bacteriën vaster aan gehecht zaten. Dit

komt waarschijnlijk omdat deze bacteriën zich eerst moeten verlengen, voordat ze in tweeën

kunnen delen. Voor het verlengen over het biomateriaal oppervlak moeten ze een

wrijvingskracht overwinnen, die groter is naarmate de aantrekking tussen de bacterie en het

biomateriaal oppervlak groter is.

Deze theorie is verder onderzocht in Hoofdstuk 4, waar de hechting en oppervlakte

groei van twee staafvormige (P. aeruginosa en Escherichia coli) en twee bolvormige (S.

epidermidis en Staphylococcus aureus) bacteriesoorten werd gemeten op drie biomateriaal

oppervlakken met een verschillende elektrische lading, sterk negatief, negatief en positief.

Omdat bacteriën een negatieve elektrische lading bezitten, worden ze elektrisch afgestoten

door negatief geladen oppervlakken, en aangetrokken door positief geladen oppervlakken. Dit

was dan ook duidelijk te zien aan de snelheid waarmee de bacteriën hechtten, die het laagst

was op het meest negatief geladen oppervlak, en het hoogst op het positief geladen oppervlak.

Oppervlakte groei van de staafvormige bacteriën was geheel afwezig op het positief geladen

materiaal, terwijl deze normaal groeiden op de negatief geladen oppervlakken. De groei van

de bolvormige bacteriën werd nauwelijks beïnvloed door de lading van het oppervlak waar ze

op gehecht waren. Op het positief geladen oppervlak worden de staafvormige bacteriën dus

waarschijnlijk zo sterk aangetrokken dat ze de wrijvingskracht niet meer kunnen overwinnen,

en niet in tweeën kunnen delen.

Biomaterialen met een positief geladen oppervlak worden dus mogelijk minder gauw

door staafvormige bacteriën geïnfecteerd, maar de omstandigheden in deze laboratorium

experimenten zijn anders dan die in het menselijk lichaam. Daarom werden schijfjes van de

biomaterialen met een verschillende oppervlakte lading en met de twee soorten staafvormige

bacteriën uit hoofdstuk 4 op het biomateriaal oppervlak geïmplanteerd in ratten (Hoofdstuk

5). Met één bacteriesoort (E. coli) bleken na twee dagen op alle negatief geladen biomateriaal

schijfjes levende bacteriën te zitten, terwijl deze maar op de helft van de positief geladen

biomateriaal schijfjes zaten. De andere bacteriesoort (P. aeruginosa) werd echter wel op alle

Samenvatting

117

biomateriaal schijfjes teruggevonden. Dit komt waarschijnlijk omdat deze bacterie veel meer

slijm kan aanmaken dan de eerste soort (E. coli). Onder deze omstandigheden kan de P.

aeruginosa bacterie misschien een slijmlaag tussen zichzelf en het geladen biomateriaal

oppervlak maken, waardoor hij geen last meer heeft van de aantrekkende kracht van het

positief geladen biomateriaal oppervlak. Deze biomaterialen met positieve oppervlakte lading

lijken dus in het lichaam biomateriaal-gerelateerde infecties te kunnen voorkomen, maar

slechts voor een beperkt aantal bacterie soorten.

Voordat de biomateriaal schijfjes werden geïmplanteerd bleek verder, dat op de

positief geladen schijfjes evenveel levende bacteriën zaten als op de negatief geladen

materialen, terwijl aan het eerste materiaal twee keer zoveel bacteriën hechtten. Blijkbaar was

ongeveer de helft van de gehechte bacteriën dood op het positief geladen materiaal. Dit komt

waarschijnlijk omdat de positieve lading van het oppervlak wordt veroorzaakt door positief

geladen stikstof atomen die in het materiaal ingebouwd zitten. In water oplosbare stoffen met

positief geladen stikstof atomen zijn bacteriedodend, omdat ze gaten kunnen maken in het

celmembraan, de ‘huid’, van de bacteriën, en worden gebruikt in desinfectie middelen. Uit

ons onderzoek blijkt dat ook biomaterialen met gebonden positief geladen stikstof atomen

bacteriedodend kunnen zijn voor gehechte bacteriën, waarschijnlijk omdat deze op dezelfde

manier ‘lek geprikt’ worden.

Het ontstaan van biomateriaal-gerelateerde infecties is niet geheel bekend. Het zou

kunnen dat bacteriën al tijdens de operatie aan het implantaat hechten. Implantaten zouden

echter ook geïnfecteerd kunnen worden in een veel later stadium, door bacteriën die in de

bloed terechtgekomen zijn. Bacteriën kunnen onder andere in de bloedbaan terechtkomen

door geïnfecteerde wondjes, ingrepen van de tandarts of blaasontsteking. Ook kunnen

bacteriën die normaal gesproken in de darmen leven soms de darmwand passeren naar de

bloedbaan. In Hoofdstuk 6 werd in gezonde ratten onderzocht of biomateriaal schijfjes 4

weken na implantatie door bacteriën in de bloedbaan geïnfecteerd kunnen worden. Hiervoor

werden óf bacteriën in de bloedbaan van de ratten geïnjecteerd, óf werd het passeren van de

darmwand door darmbacteriën van de ratten gestimuleerd. Het bleek dat de biomateriaal

schijfjes niet via het bloed geïnfecteerd konden worden. Echter, 5% van de geïmplanteerde

schijfjes werd geïnfecteerd door onbekende bacteriën die waarschijnlijk tijdens de implantatie

op het materiaal waren gekomen. Waarschijnlijk is ook bij mensen met een gezond afweer

systeem de kans klein, dat implantaten in een later stadium na implantatie alsnog door

bacteriën in de bloedbaan worden geïnfecteerd.

Samenvatting

118

Nu dat we wisten, dat ook gebonden positieve lading bacteriën ‘lek’ kan prikken,

hebben we positief geladen stikstof atomen met een lange moleculaire staart, een soort

prikker, gebruikt, in de hoop dat dit lek prikken beter gaat. Deze positief geladen stikstof

atomen hebben we chemisch gebonden aan siliconen rubber, een veel gebruikt biomateriaal

(Hoofdstuk 7). Vervolgens hebben we onze vier soorten bacteriën anderhalf uur gehecht aan

de materialen en daarna gekeken of ze levend waren. Het bleek dat vrijwel alle bolvormige

bacteriën (S. aureus en S. epidermidis) niet meer in leven waren op het behandelde siliconen

rubber, terwijl de meeste gehechte bacteriën op onbehandeld siliconen rubber wel leefden.

Ook veel (ca. 75%) van de gehechte staafvormige bacteriën (P. aeruginosa en E. coli) waren

niet meer in leven op het behandelde siliconen rubber, terwijl dit niet het geval was op het

onbehandelde siliconen rubber. De effectiviteit van de bacterie dodende laag werd ook hier

getest door implantatie van behandelde en onbehandelde schijfjes siliconen rubber in ratten

met één van de bacteriën (S. aureus) erop gebracht. Wanneer de bacteriën voor implantatie

werden gehecht aan de materialen werden na 3 en 7 dagen nagenoeg geen levende bacteriën

gevonden op de behandelde siliconen rubber schijfjes, terwijl deze wel aanwezig waren op

vrijwel alle onbehandelde siliconen rubber schijfjes. Wanneer de bacteriën na de implantatie

in de buurt van de schijfjes werden geïnjecteerd werden ook op de meeste behandelde

siliconen rubber schijfjes levende bacteriën terug gevonden, maar veel minder dan op de

onbehandelde materialen.

We kunnen uit het onderzoek concluderen dat het positief maken van het biomateriaal

oppervlak met positief geladen stikstof atomen met een lange moleculaire staart biomateriaal-

gerelateerde infecties kan voorkomen. De oppervlakte groei van de onderzochte bolvormige

bacteriën wordt hierop effectief voorkomen door het doden van deze bacteriën. Ook een

gedeelte van de onderzochte staafvormige bacteriën wordt op zo’n oppervlak gedood, en de

overlevende bacteriën kunnen moeilijk groeien op het positief geladen oppervlak door de

sterke aantrekkingskracht ervan. Deze positief geladen stikstof atomen kunnen in de toekomst

chemisch worden gebonden aan biomedische implantaten en hulpmiddelen, zoals

bijvoorbeeld heup protheses, kunst hartkleppen of urineweg katheters, waardoor infecties, en

de daarmee gepaard gaande ernstige problemen voor de patiënt, voorkomen kunnen worden.

119

Nawoord

Ruim vier jaar geleden vertrok ik naar het hoge noorden om aan dit AIO project te beginnen.

Het werd een leuke en leerzame tijd, en uiteindelijk ontstond zelfs dit proefschrift. Dit was

natuurlijk niet gelukt zonder hulp van veel mensen, die ik hier van harte wil bedanken.

Henk Busscher, jouw betrokken, enthousiaste en heldere begeleiding heb ik altijd erg

prettig gevonden. Ook als het onderzoek maar niet wilde opschieten hebben je optimisme en

vertrouwen me gesteund.

Henny van der Mei, telkens weer stond ik er versteld van hoe gemakkelijk jij een

praktisch probleem op kon lossen. Van onze gezamenlijke congresbezoekjes heb ik genoten.

Paul Nieuwenhuis, als tweede promotor kwam jij altijd met nieuwe gezichtspunten en

hypotheses. Dankzij jou heb ik veel dierexperimenteel werk gedaan, wat het proefschrift

interessanter heeft gemaakt.

Flip Klatter, je hebt me geweldig geholpen met mijn ratten. Zonder jou had ik niet

geweten wat ik met de beestjes aan moest.

Jan Feijen, u heeft een belangrijke richting aan dit proefschrift gegeven. Bedankt voor

de mogelijkheid op de TU Twente onderzoek te doen, en dat u mijn derde promotor wilde

zijn.

Dirk Grijpma, je hebt me bij jullie op de TU Twente wegwijs gemaakt, en de kneepjes

van de polymeerchemie bijgebracht.

Iedereen bij de BME bedankt voor zijn bijdrage, met name Ellen van Drooge en Ina

Heidema-Kol voor de administratieve ondersteuning en hulp bij de layout, Joop de Vries voor

de ESCA metingen en niet te vergeten Betsy van de Belt-Gritter en anderen voor het gieten

van de vele bloedplaten.

De medewerkers van het Centraal Dieren Laboratorium wil ik bedanken voor hun

technische inbreng en de goede verzorging van de dieren.

De mannen van de instrumentmakerij, bedankt voor alle onderdelen die jullie gemaakt

en gerepareerd hebben.

Prof. Dankert, Prof. Degener en Prof. Schouten ben ik erkentelijk voor de snelle

beoordeling van het manuscript.

Dankzij alle collega’s was de werksfeer altijd erg goed, wat het meest bijdroeg aan het

plezier in mijn werk. Albert Poortinga en Gerda Bruinsma, leuk dat jullie paranimf willen

120

zijn, en net als mijn andere kamergenoten door de jaren heen, Chris van Hoogmoed, Dewi

Bakker en Virginia Vadillo Rodríguez bedankt voor de gezelligheid.

Aletta van Rheede, je hebt een leuke omslag gemaakt, bedankt.

Tamara, bedankt dat je me achterna kwam naar Groningen, ik vond het een leuke tijd.

Verder heb je me goed geholpen met de niet wetenschappelijke inhoud van dit proefschrift.

Mijn familie en vrienden, jullie belangstelling en medeleven heb ik erg gewaardeerd.

Mama, dit boekje is voor jou. Jammer dat je er niet meer bij kunt zijn.

Bart.

Positively charged biomaterials exert antimicrobial effects on gram-negative bacilli in rats

Documents

Transcript of Positively charged biomaterials exert antimicrobial effects on gram-negative bacilli in rats