pathogenic factors of pseudomonas cepacia isolates from patients with cystic fibrosis

J. Med. Microbiol. - Vol. 33 (1990), 115-120 0 1990 The Pathological Society of Great Britain and Ireland

0022-261 5/90/0033-011 5/$10.00

Pathogenic factors of Pseudomonas cepacia isolates from patients with cystic fibrosis

A. R. GESSNER and J. E. MORTENSEN’

Department of Laboratories, St Christopher’s Hospital for Children, Fifth and Lehigh Avenue, Philadelphia and Department of Pediatrics, Temple University School of Medicine, 340 7 North Broad Street, Philadelphia, PA

79 7 33, USA

Summary. One hundred and nineteen isolates of Pseudomonas cepacia, 98 of which were from cystic fibrosis (CF) patients and 21 from environmental and other human sources, were examined for biochemical and exo-enzymatic properties that may contribute to the pathogenicity of this bacterium. The following characteristics were demonstrated significantly more frequently in isolates from CF patients than in control isolates : production of catalase, ornithine decarboxylase, valine aminopeptidase, C14 lipase, alginase and trypsin; reduction of nitrate to nitrite; hydrolysis of urea and xanthine ; complete haemolysis on bovine red blood cells; cold-sensitive haemolysis on human red blood cells; greening of horse and rabbit red blood cells. The role of these factors in the pulmonary disease associated with cystic fibrosis is not clear. However, several factors which have been reported previously as being associated with pathogenic processes with other bacteria have now been described in P. cepacia. Additional factors not previously reported as “pathogenicity factors” are also described.

Introduction

Pseudomonas cepacia has become an important pathogen in patients with cystic fibrosis (CF). In 1980, Rosenstein and Hall reported a case of pneumonia and septicaemia caused by P. cepacia in a 17-year-old CF patient. This case is of interest because it typifies the syndrome of rapid pulmonary deterioration frequently accompanied by bacterae- mia in CF patients colonised by P . cepacia. Following this case report, other CF centres evaluated their experiences with P . cepacia colonisation. In Toronto, patient colonisation rates with this organism had increased from 10% in 1971 to 18% in 1981 while the rate of colonisation with P . aeruginosa had remained constant over the same period.2 In the same study, the presence of P . cepacia correlated with a greater impairment of pulmonary function than did P. aeruginosa colonisation. A CF centre in Cleveland demonstrated a doubling in P . cepacia colonisation from 1978 to

Received 23 Nov. 1989; accepted with amendments 14 April 1990. *Correspondence should be sent to Dr J. E. Mortensen, Director, Microbiology Laboratory, St Christopher’s Hospital for Children, Fifth and Lehigh Avenue, Philadelphia, PA 19133, USA.

1983 and the proportion of patients from which P . cepacia was isolated at autopsy increased sixfold.

Certain characteristics of P . cepacia may explain the frequency of recovery from CF patients. P. cepacia has minimal growth requirements and can survive in hostile environments such as distilled water and disinfectants for extended periods.4* The organism has been associated with nosocomial outbreaks6.’ and has been recovered from pulmonary equipment.6

The virulence of P . cepacia in animals has been examined in two studies. In one, P . cepacia recovered from contaminated disinfectant was lethal to mice only at doses > lo9 cfulml.’ In a second study, a P . cepacia isolate from a CF patient was found to colonise burned skin in a mouse infection model for at least 3 weeks without tissue invasion or mortality. In this same model, P. aeruginosa was invasive and caused deaths within 2 days.*

The finding of long term survival in hostile environments and the ability to colonise damaged tissue may explain the acquisition and persistence of P . cepacia in CF patients, but it does not explain the role played by this organism in the pulmonary disease of these patients. In the present study we examined two populations of P . cepacia-isolates

115

116 A. R. GESSNER AND J. E. MORTENSEN

from patients with CF and from other sources-for the production of a wide variety of enzymes and potential pathogenicity factors that may play a role in the pulmonary disease associated with P . cepacia in CF patients.

Materials and methods

Strains, test conditions, and reagents

One hundred and nineteen P. cepacia isolates were studied. Of these, 98 were respiratory isolates from CF patients-85 from patients seen at the Cystic Fibrosis Center, St Christopher’s Hospital for Children and 13 isolates from other CF centres in Cleveland, OH, USA (3), Toronto, Canada (3), Seattle, WA, USA (3)’ Calgary, Ontario, Canada (2), Columbia University (1) and Yale- New Haven, CT, USA (1). Primary identification of each CF isolate from St Christopher’s Hospital for Children was performed in the clinical microbiology laboratory. These isolates were then frozen in litmus milk and held at - 70°C. The other 21 isolates were from environmental sources and other human infections (control isolates). The number and source of this subset of organisms were as follows : American Type Culture Collection (ATCC) environmental (lo), ATCC human infection (7)’ Phila- delphia environmental (3) and Philadelphia human infection (1).

Exo-enzymatic plate assays and biochemical tests were incubated at 37°C for 48 h, then held at 25°C for another 5 days. Biochemical results were recorded at 24 and 48 h and 7 days, and exo-enzymatic results were recorded at 48 and 72 h and 7 days, unless otherwise stated. The API- ZYM panel (Analytab Products, Inc., Plainview, NY, USA) was also used to screen for specific enzymatic activity. Quality control was performed on all assays with known positive and negative organisms.

Media for these assays was purchased from Difco (Detroit, MI, USA), BBL (Cockeysville, MD, USA), Sacks Farms (Evans City, PA, USA), Remel (Lenexa, KA, USA), Sigma Chemical Co. (St Louis, MO, USA), Flow Laboratories, Inc. (McLean, VA, USA), Scott Laboratories, Inc. (Carson, CA, USA), and Analytab Products (Plainview, NY, USA). All forms of media were prepared and used according to the manufacturers’ instructions.

Biochemical tests

Biochemical tests were performed in accordance with established methods.g. lo The tests included : citrate utilisation on Simmons citrate agar, growth on cetrimide agar, growth at 42”C, growth on PC agar (Scott), nitrate reduction, urea hydrolysis, aesculin hydrolysis, production of catalase and oxidase. Catalase and oxidase activities were evaluated after incubation at 37°C for 48 h on heated blood agar plates. Nitrate reduction was tested after incubation for 48 h at 37°C.

Tests with amino acids and nucleic acids

Determination of amino-acid decarboxylation or di- hydrolation was performed by Greenwood’s method for non-fermentative organisms. ’ The amino acids tested were arginine, ornithine and lysine. The organisms were grown at 37°C in broth containing glucose 0.5%, KH2P0, 5% and amino acid 0.5%. After 24 h, one drop of 1 0 ~ NaOH was added followed by 1 ml of ninhydrin (Sigma N-4876) 0.1% in chloroform. Development of a purple colour in the chloroform layer at 5 min was positive. Phenylalanine deaminase activity was determined after incubation at 37°C for 24 h by flooding the medium with ferric chloride. Leucine, valine and cystine aminopepti- dases were tested with the API-ZYM panel and were elevated according to manufacturer’s instructions. Activ- ity in xanthine and tyrosine were assayed by clearing of the media (Remel) around the inoculum. DNAase activity was demonstrated by a zone of clearing around the inoculum when the DNAase plate (Difco) was flooded with 1~ HC1 after incubation for 72 h.” RNAase test agar contained RNA (Sigma R-7 125 from Baker’s Yeast) 0.2% in Brain Heart Infusion (BHI) agar base. A test was considered positive if, after incubation at 37°C for 5 days, flooding with 1~ HC1 produced a zone of clearing.

Tests with lipids

Several methods were used to screen for lipolytic activity. Lecithinase activity was detected by growth on (Remel) BHI agar with egg yolk 10%.l2 Esterase was detected by growth on BHI agar containing Tween 80 (Sigma P-1754) 1 %. ’ C4 esterase, C8 esterase/lipase, C14 lipase, acid and alkaline phosphatases and phosphohydrolase were detected in the API-ZYM panel. l4

Haemolysin

Haemolytic activity was determined on five types of blood agar: sheep (BBL), bovine, rabbit, horse and human (Sacks Farms). Plates were incubated at 37°C for 48 h and then at 0°C for a further 24 h. Results were recorded at 24,48 and 72 h and reported as no haemolysis, greening of the media or complete haemolysis. ’

Tests with proteins and polysaccharides

Enzymatic activity against cartilage (Sigma C-5268)’ collagen (Sigma C-9879) and elastin (Sigma E-1625) was determined by incorporation of substrate 2% in BHI agar. l 5 Caseinase activity was determined on BHI with skimmed milk (Remel, casein) 50% and agar l % . 1 5 Albuminase was detected by growth on BHI agar with Bovine Serum Albumin (Sigma A-4503) 8%. Gelatinase activity was evaluated by liquefaction of gelatin (Difco) 12% in BHI broth.g Digestion of algin was observed in the algin-charcoal media of von Riesen16 (yeast extract, Difco, 0.5%; Algin, Sigma A-7 128, 1%; bromothymol blue 0.002%; charcoal 0.1%). Amylase was detected in BHI agar with starch 2%.’ Mucinase activity was tested

PATHOGENICITY OF P. CEPACIA IN CYSTIC FIBROSIS 117

on BHI agar with hog gastric mucin (Sigma M-2378) 0.28%.13 After incubation for 7 days, the medium was flooded with calcium chloride 1 % ; mucinase-positive strains produced clear zones around the inoculum.

The other enzymes detected in the API-ZYM panel were : trypsin, chymotrypsin, a- and p-galactosidases, p- glucuronidase, a- and P-glucosidases, a-mannosidase, a- fucosidase and N-acetyl-b-glucosaminidase. ’

Cy to toxicity

Cytotoxicity was evaluated by the method of McKevitt and Woods” in which CHO cells and Vero cells were grown to monolayers in RPMI 1640 medium (Flow Laboratories, Inc.) supplemented with fetal calf serum 10% and gentamicin 0.08%. The cells were resuspended in fresh growth medium and incubated for 24 h at 37°C in COz 5%. Filter-sterilised supernate (100 pl) from P . cepacia cultures grown in RPMI 1640 for 48 h at 37°C were pipetted on to the cells, together with 100 pl of fresh growth medium. Cells were observed for cytopathic effects after 24 and 48 h and 5 days at 37°C in COz 5%.

Statistical analysis

Statistical analysis of all results was performed with Fisher’s Exact test to compare the proportion of positive tests in each group. Significance was indicated by p values 50.05.

Results

Test results recorded as positive on or before the final reading were considered positive and are presented as such in the following sections and tables.

Biochemical properties

The biochemical properties of P . cepacia are summarised in table I. In these biochemical tests, the CF isolates gave positive results for 290% of the isolates for four tests-catalase production, growth at 42”C, growth on PC agar, and citrate utilisation. With the control isolates, 2 90% gave positive results in two tests-citrate utilisation and urea hydrolysis. Statistically significant differences between CF isolates and controls were noted for catalase production, urea hydrolysis and nitrate reduction.

Tests with amino acids and nucleic acids

The results of the amino acid and nucleic acid assays are summarised in table 11. The CF and control isolates gave positive results with 2 90% of isolates only for tyrosine and leucine aminopepti-

Table I. Comparison of biochemical properties of P . cepacia isolates from CF patients and control isolates

Property

Number (“A) of isolates that gave positive results

CF Control (98) (21) p value

Catalase production Oxidase production Citrate utilisation Growth on cetrimide agar Growth at 42°C Growth on PC agar Nitrate reduction

NO2 N 2 Aesculin hydrolysis

Urea hydrolysis

98 (1 00)

92 (94) 81 (83) 89 (91) 92 (94)

8 (8) l (1)

t62 (64)

10 (10)

*63 (89)

18 (86) 1(5)

20 (95) 16 (76) 17 (81) 18 (86)

8 (38) 0 (0)

17 (81) 21 (100)

0.005 NS NS NS NS NS

0.00 1 NS NS

0*0003

* 71 isolates tested. 97 isolates tested.

NS, not significant.

Table 11. Comparison of amino acid and nucleic acid degradation by CF and control isolates of P . cepacia

Property

Number (%) of isolates that gave positive results

CF Control (98) (21) pvalue

Lysine decarboxylase Ornithine decarboxylase Xanthine hydrolysis Tyrosine hydrolysis RNAase Leucine aminopeptidase Valine aminopeptidase Cystine aminopeptidase

81 (83) 18 (86) 78 (80) 6(29) 14 (14) 17 (81) 94 (96) 20 (95) 79 (81) 19 (90) 94 (96) 21 (100) 4 (4) 4 (19)

lO(10) O(0)

NS < 0.001 < 0.00 1

NS NS NS

0.032 NS

None gave positive results in tests for arginine dihydrolase, DNAase and phenylalanine deaminase.

dases. With the control isolates, 290%. gave positive results for RNAase. Statistically significant differences were noted with ornithine decarboxylase, xanthine hydrolysis and valine aminopeptidase .

Tests with lipids

The results of the lipid assays are shown in table 111. With isolates of P . cepacia from both CF patients and controls, 2 90% gave positive results


Table 111. Comparison of lipolytic activity of isolates of P . cepacia from CF patients and controls

Property



Lipase (Tween 80) Lecithinase (egg yolk) C4 esterase C8 esterase/lipase C 14 lipase Acid phosphatase Alkaline phosphatase Phosphohydrolase

the red blood cells was detected in 14-25% of the CF isolates and &lo% of the control isolates. Statistically significant differences were noted for complete haemolysis on bovine red blood cells, cold-sensitive haemolysis on human red blood cells and greening on horse and rabbit red blood cells. All haemolytic activity is shown in table IV.

74 (76) 18 (86) NS 68(69) 17(81) NS 95(97) 21 (100) NS 98(100) 21 (100) NS 35(36) 14(67) 0.009 96(98) 21 (100) NS 86(88) 19(91) NS 89(91) 21 (100) NS

Tests with proteins and polysaccharides

None of the tests for proteolytic or saccharolytic enzymes included in this study demonstrated 2 90% positive results with CF or control isolates. All results are shown in table V. Statistically significant differences were found for alginase and trypsin.

for the C4 and C8 esterases, acid phosphatase and phosphohydrolase. With the control isolates, 2 90% also gave positive results for alkaline phosphatase. A statistically significant difference was seen for the C14 lipase only.

Haemolysis

Some of the P. cepacia strains had an effect on some or all of the five types of red blood cells. This activity was shown as complete lysis of the red blood cells, cold-dependent lysis and greening of the red blood cells. Complete, non-temperature- dependent lysis was exhibited in 3-6% of the cystic isolates and 5-20% of the control isolates. Cold- sensitive haemolysis was noted for 2-18% of CF isolates and 0-20% of control isolates. Greening of

Cy to toxicity

any cytotoxicity for CHO or Vero cells. None of the CF nor the control isolates showed

Discussion

P . cepacia colonisation in CF patients has been identified as a potentially significant factor in the pulmonary disease process in these patients. 1-3 One explanation for this association is that this organism may play an active role through proteolytic, saccharolytic or lipolytic activity. Such factors have been well characterised in P . aeruginosa." The proteases and elastases of P. aeruginosa have been shown to injure pulmonary tissue directly." The report by Marks' cites that "intratracheal admin- istrations of highly purified P . aeruginosa proteases

Table IV. Comparison of haemolytic activity by 97 CF and 20 control isolates of P . cepacia against various red blood cell species

Cell species

Horse Human Bovine Rabbit Sheep

Number (%) of isolates that produced

cold-sensitive complete haemolysis haemolysis greening

CF control CF control CF control

Percentages may exceed 100% due to the expression of multiple haemolytic activities. Significant differences (p < 0.05) : lysis of bovine erythrocytes (p = 0.016), cold sensitive lysis of human erythrocytes (p = 0-030), greening of horse (p = 0.006) and rabbit (p = 0.010) erythrocytes.

PATHOGENICITY OF P. CEPACZA IN CYSTIC FIBROSIS 119

Table V. Comparison of proteolytic and saccharolytic activity by CF and control isolates of P. cepacia

Property

Caseinase Gelatinase Alginase Mucinase Albuminase Trypsin C h ymo try psi n a-Glucosidase /3-Glucosidase N-acetyl-

a-Fucosidase /?-glucosarninidase



79 (81) 15 (71) 38 (39) 1 1 (52) 16 (16) 10 (48) 9 (9) 3 (14) 5 (5) 1 (5) 4 (4) 5 (24) 3 (3) 0 (0) 0 (0) 1 (5) 7 (7) 4 (19)

41 (42) 13 (62) 0 (0) l (5 )

~

NS NS

0.003 NS NS

0.009 NS NS NS

NS NS

None of the isolates gave positive results in tests for a- mannosidase, a- and /3-galactosidase, /3-glucuronidase, cartila- ginase, collagenase, elastase and amylase.

in rabbits elicits extensive, grossly observable lung damage”. Also noted by Marks was similar pulmonary toxicity by elastase following intranasal ad- ministration in mice.

A few investigators have studied bacterial products that may be associated with the pathogenicity of this organism ; however, no clear associations have been reported.*. 1 7 7 l 9 Therefore, this project was undertaken to establish the extracellular factors of P . cepacia that may be involved in the pulmonary disease process of CF and to delineate areas for further investigation.

Recently, Lonon and associates have reported experiments on the lipolytic activity of P . cepacia in tests with egg yolk agar, four different Tweens (20,40,60 and 80) and a chromogenic substrate for phospholipase C. These researchers detected no phospholipase C activity with egg yolk, but activity was observed with the other substrates. The present study supports and expands on their findings.

Nakazawa and associates have recently reported the characterisation of a haemolysin produced by P. cepacia with activity against sheep erythrocytes.” They showed that c. 4% of the P . cepacia strains tested produced complete haemolysis of sheep red blood cells, which is consistent with our findings. We examined erythrocytes from five sources and found that the percentages of isolates

haemolytic for bovine, rabbit, human and horse red blood cells differed from the percentages found with sheep red blood cells. It is likely that differences in the composition of the cell-membranes of the different erythrocytes account for the different activities. Furthermore, when the haemolytic activity of P . cepacia was examined for related phenomena, a larger number of isolates demonstrated either greening of the red blood cells or cold-sensitive haemolysis (see table IV). The differences in host range as well as the complex combination of phenomena make it especially difficult to evaluate the role of haemolytic activity in the pathogenicity of P . cepacia.

The apparent lack of activity for a number of substrates does not necessarily show that the organism does not possess the appropriate enzymes. One explanation for the apparent lack of proteolytic or saccharolytic activity maybe the availability of iron. Virulence of P . aeruginosa in animal models, as well as amounts of extracellular enzymes produced in culture, have been shown to be affected by the concentration of iron in the growth media.20-22

Consistent with the report by McKevitt and Woods, none of the P . cepacia isolates was cyto- toxic. ’’

In addition to a number of potentially pathogenic factors produced by all strains of P . cepacia, we found several significant differences in the expression of biochemical reactions or the production of exo-enzymes by P . cepacia isolates from CF patients when compared to control isolates. These included : production of catalase, ornithine decarboxylase, valine aminopeptidase, C 14 lipase, alginase and trypsin; reduction of nitrate to nitrite; hydrolysis of urea and xanthine; complete haemolysis on bovine red blood cells; cold sensitive haemolysis on human red blood cells; and greening of horse and rabbit red blood cells.

The role of any of these factors in the disease associated with respiratory colonisation or infection in CF patients is not clear, but the present study has provided a number of avenues for future research. Using a variation on the chronic respiratory infection model in rats developed by Cash et aZ.,23 the role of specific biochemical reactions or enzyme production in the pathogenesis of chronic infection with P . cepacia may be determined.

We thank the following for their contributions : Drs LiPuma and Stull from the Medical College of Pennsylvania and Dr A. Kuritza for P. cepuciu and Dr Janet Clark from St Christopher’s Hospital for Children for CHO cells.


REFERENCES

1. Rosenstein BJ, Hall DE. Pneumonia and septicemia due to Pseudomonas cepacia in a patient with cystic fibrosis. John Hopkins Med J 1980; 147: 188-189.

2. Isles A, Maclusky I, Corey M et al. Pseudomonas cepacia infection in cystic fibrosis patients: an emerging problem. J Pediatr 1984; 104: 206-210.

3. Thomassen MJ, Demko CA, Klinger JD, Stern RC. Pseudomonas cepacia colonization among patients with cystic fibrosis. A new opportunist. A m Rev Respir Dis

4. Carson LA, Favero MS, Bond WW, Petersen NJ. Morphol- ogical, biochemical, and growth characteristics of Pseudomonas cepacia from distilled water. Appl Micro- bioll973; 25: 476-483.

5. Geftic SG, Heymann H, Adair FW. Fourteen-year survival of Pseudomonas cepacia in a salts solution preserved with benzalkonium chloride. Appl Environ Microbiol

6. Gelbart SM, Reinhardt GF, Greenlee HB. Pseudomonas cepacia strains isolated from water reservoirs of un- heated nebulizers. J Clin Microbioll976; 3: 62-66.

7. Sobel JD, Hashman N, Reinherz G, Merzbach D. Nosocom- ial Pseudomonas cepacia infection associated with chlorhexidine contamination. Am JMed 1982; 73: 183- 186.

8. Stover GB, Drake DR, Montie TC. Virulence of different Pseudomonas species in a burned mouse model: tissue colonization by Pseudomonas cepacia. Infect Immun

9. McFaddin JF. Biochemical tests for identification of medical bacteria, 2nd edn. Baltimore, MD, Williams and Wilkins. 1980: 4-12,51-63,94-113,236245,249- 260,269-274 and 298-308.

10. Phillips E, Nash P. Culture media. In: Lennette EH, Balows A, Hausler WJ, Shadomy HJ (eds) Manual of clinical microbiology, 4th edn. Washington, DC, American Society for Microbiology. 1985 : 1051-1092.

11. Greenwood JR. Methods of isolation and identification of glucose-nonfermenting gram-negative rods. In : Gilardi GL (ed) Nonfermentative gram-negative rods : labora-

1985; 131: 791-796.

1979; 37: 505-510.

1983; 41: 1099-1 104.

tory identification and clinical aspects (Microbiology Series, vol16) New York, Marcel Dekker Inc. 1985: 1- 16.

12. Lonon MK, Woods DE, Straus DC. Production of lipase by clinical isolates of Pseudomonas cepacia. J CIin Microbiol

13. O’Brien M, Davis GHG. Enzymatic profile of Pseudomonas maltophilia. J Clin Microbioll982; 16: 417-421.

14. Poh CL, Loh GK. Enzymatic characterization of Pseudo- monas cepacia by API-ZYM profile. J Clin Microbiol 1988; 26: 607-608.

15. Janda JM, Bottone EJ. Pathogenic properties and enzyme profiles of Pseudomonas and Acinetobacter. In : Gilardi GL (ed) Nonfermentative gram-negative rods : laboratory identification and clinical aspects (Microbiology Series, vol 16) New York, Marcel Dekker Inc. 1985:

16. Von Riesen VL. Digestion of algin by Pseudornonas maltophilia and Pseudomonas putida. Appl Environ Microbioll980; 39: 92-96.

17. McKevitt AI, Woods DE. Characterization of Pseudomonas cepacia isolates from patients with cystic fibrosis. J Clin Microbioll984; 19: 291-293.

18. Marks MI. The pathogenesis and treatment of pulmonary infections in patients with cystic fibrosis. J Pediatr

19. Nakazawa T, Yamade Y, Ishibashi M. Characterization of hemolysin in extracellular products of Pseudomonas cepacia. JClin MicrobioIl987; 25: 195-198.

20. Bjorn MJ, Sokol PA, Iglewski BH. Influence of iron on yields of extracellular products in Pseudomonas aeruginosa cultures. J Bacterioll979; 138: 193-200.

21. Sokol PA, Woods DE. Relationship of iron and extracellular virulence factors to Pseudomonas aeruginosa lung infections. J M e d Microbioll984; 18: 125-133.

22. Woods DE, Cryz SJ, Friedman RL, Iglewski BH. Contri- bution of toxin A and elastase to virulence of Pseudo- monas aeruginosa in chronic lung infections of rats. Infect Immun 1982; 36: 1223-1228.

23. Cash HA, Woods DE, McCullough B, Johanson WG, Bass JA. A rat model of chronic respiratory infection with Pseudomonas aeruginosa. Am Rev Respir Dis 1979 ; 119 : 453459.

1988 ; 26 : 979-984.

233-254.

1981; 98: 173-179.

pathogenic factors of pseudomonas cepacia isolates from patients with cystic fibrosis

Documents