High prevalence of phagocytic-resistant capsular serotypes of Klebsiella pneumoniae in liver abscess

Microbes and Infection 6 (2004) 1191–1198 www.elsevier.com/locate/micinf

Original article

High prevalence of phagocytic-resistant capsular serotypes of Klebsiella pneumoniae in liver abscess Jung-Chung Lin a,b, Feng-Yee Chang a, Chang-Phone Fung c, Jin-Zhen Xu d, Hsiao-Pei Cheng a, Jaang-Jiun Wang e, Li-Yueh Huang f, L.K. Siu f,* a

Division of Infectious Diseases and Tropical Medicine, Department of Internal Medicine, Tri-Service General Hospital, Taipei, Taiwan b Graduate Institute of Medical Sciences, National Defense Medical Center, Taipei, Taiwan c Section of Infectious Diseases, Department of Medicine, Taipei Veterans General Hospital, National Yang-Ming University, Taipei, Taiwan d Division of Biostatistics, National Health Research Institutes, Taipei, Taiwan e Department and Institute of Biology and Anatomy, National Defense Medical Center, Taipei, Taiwan f Division of Clinical Research, National Health Research Institutes (99), 128, Yen-Chiu-Yuan Road, Sec. 2, Taipei 11529, Taiwan Received 19 April 2004; accepted 17 June 2004 Available online 15 September 2004

Abstract To better understand the role of capsular polysaccharide (CPS) K1 or K2 in Klebsiella pneumoniae liver abscess as well as the development of metastasis to eye, neutrophil phagocytosis of 70 CPS isolates including K1 (n = 23)/K2 (n = 10), non-K1/K2 (n = 37) was evaluated by flow cytometry, fluorescence imaging, and electron microscopy. K1/K2 isolates were significantly more resistant to phagocytosis (P < 0.0001) than non-K1/K2 isolates and displayed increased resistance to intracellular killing. Although mucoid phenotype (M-type) K1/K2 isolates were significantly more resistant to phagocytosis (P = 0.0029) than M-type non-K1/K2, no significant difference in the phagocytosis rate was observed between K1/K2 isolates with M-type and non-M-type (P = 0.0924). Mucoidy is an associated factor that was predominant in K1/K2 isolates, but which itself is not an independent influence on phagocytic resistance. The K1/K2 CPS proved significantly more resistant to phagocytosis than non-K1/K2 CPS in liver abscess isolates (P < 0.0001) and non-abscess isolates (P = 0.0001), suggesting that K1/K2 isolates were generally more virulent in both liver abscess and in non-liver abscess conditions. These findings indicate that resistance of CPS K1 or K2 K. pneumoniae to phagocytosis and intracellular killing presumably contributes to their high prevalence in liver abscess and uniquely in endophthalmitis. © 2004 Elsevier SAS. All rights reserved. Keywords: Klebsiella pneumoniae; Liver abscess; Serotype K1; Phagocytosis

1. Introduction Klebsiella pneumoniae is the most prominent member of this genus in Klebsiella-related infections as well as more generally in Gram-negative sepsis [1,2]. The recent emergence and dissemination of extended spectrum b-lactamases has further complicated treatment of these infections [1,2]. Although K. pneumoniae is a common cause of nosocomial infections [3], community-acquired infections are specific in some instances [4]. Infections are often associated with high morbidity and mortality [5].

* Corresponding author. Tél.: +886-2-2652-4094; Fax: +886-2-2789-0254. E-mail address: [email protected] (L.K. Siu). 1286-4579/$ - see front matter © 2004 Elsevier SAS. All rights reserved. doi:10.1016/j.micinf.2004.06.003

K. pneumoniae is the most prevalent bacteria isolated from pyogenic liver abscess in Trinidad, the United States, Singapore, Japan, Thailand, and Taiwan [6–11]. Cases have also been reported from Australia, Hong Kong, and Spain [6,12–14]. Diabetes mellitus (DM) is the major underlying condition associated with K. pneumoniae liver abscess [6,7,15–17], although the reasons for this association remain mysterious. Approximately 10% of liver abscess cases are complicated by the development of endophthalmitis [7]. An extremely poor visual outcome is common in these cases [12,13,17–19]. The relationship of the liver abscess to the host or environment, as well as bacterial pathogenesis, remains uncertain. Previously, our epidemiological surveys found a high prevalence of encapsulated strains of K. pneumoniae in liver

1192

J.-C. Lin et al. / Microbes and Infection 6 (2004) 1191–1198

abscesses, and K1 or K2 serotypes of K. pneumoniae isolates in cases complicated by endophthalmitis [4,7]. The observation that the majority of liver abscess isolates of K. pneumoniae have a well-defined capsule with K1 or K2 serotype is consistent with the possibility that the presence of a specific capsular serotype is an important risk or virulence factor for liver abscess and, in a minority of cases, for endophthalmitis. In microbial infections, polymorphonuclear leukocytes (PMNs; also called neutrophils) are critical components of the innate immune system that functions to protect the host by migrating to the site of inflammation and engulfing the invading pathogens. The rapid delivery of PMNs to sites of infection is crucial in controlling an acute infection [20]. Conversely, the antiphagocytic property of specific capsular polysaccharide (CPS) is believed to be an important virulence mechanism in infections such as those of the liver and eye. However, the role of K. pneumoniae CPS K1 or K2 in liver abscess and development of endophthalmitis is unclear. To assess this role, we used an experimental model of phagocytosis for K. pneumoniae with different CPSs isolated from patients with and without liver abscess and compared the resistance of the isolates to neutrophil-mediated phagocytosis.

2. Materials and methods 2.1. Collection of K. pneumoniae with different serotypes and sites of isolation K. pneumoniae strains with different capsular serotypes were obtained previously [4,7,21] from patients who had developed community-acquired or nosocomial infections. These strains were isolated from liver aspirate, blood, and sputum (Table 1). Serotyping was assessed by the capsular swelling technique [7]. Mucoidy was assessed by visual inspection of colonies growing on tryptic soy agar (TSA), following the inoculation of the agar and incubation at 37 °C [22]. Control K. pneumoniae serotypes including ATCC4208 (K1), ATCC13883 (K3), and ATCC700603 (K6) were acquired from the American Type Culture Collection (ATCC, Rockville, MD). 2.2. Isolation of human neutrophils The isolation of neutrophils from three healthy volunteers was approved by the Ethics Committee of the National Health Research Institutes Taipei, Taiwan. The three volunteers were screened after giving informed written consent. Neutrophils were isolated as previously described [23]. Sixty milliliter of freshly drawn, heparinized blood was mixed with an equal volume of a dextran–saline solution, and sedimentation of particulates was facilitated during 40 min at room temperature. The leukocyte-rich supernatant was layered on a density gradient constructed using Ficoll-Hypaque (Pharmacia, Taiwan) and centrifuged at 400 × g, 40 min, at 20 °C.

Table 1 Serotypes and sites of the isolation of K. pneumoniae selected in the current study Serotype

Liver aspirate 14 4

K1 K2 K3 K5 K6 K7 K8 K9 K15 K16 K17 K20 K21 K28 1 K29 K31 K32 K38 K54 K55 K57 Non-typeable 1

Number of isolates Blood Wound Sputum 4 2 1 6 1 2

Other a 1 1 1

1 1 1 1 3 1 1 1 1

2

1 1 1 1

3 5 3 1 1

1

a Control isolates with ATCC4208 (serotype K1), ATCC13883 (serotype K3) and ATCC700603 (serotype K6).

The pellet was collected and erythrocytes were removed by hypotonic lysis. After 30 s, the isotonicity was restored using hypertonic saline. The pellet was resuspended in ice-cold phosphate-buffered saline (PBS) and the cell concentration was adjusted to 1 × 107 cells per ml. Viability was over 95% as determined by trypan blue exclusion. 2.3. Fluorescence labeling of bacteria Labeling was performed as previously described [24]. K. pneumoniae isolate and control suspensions were individually incubated overnight at 37 °C. The concentration was approximated using photospectrometry (Olympus, US). The percentage of bacterial viability in an aliquot of each population was determined by quantitative plate counting. Populations were then heat-killed for 60 min in a 70 °C water bath and quantitative colony count determination of population viability was again done. The bacteria were washed with PBS and labeled with fluorescein isothiocyanate (FITC) by incubation with 0.1 mg/ml FITC (Sigma Chemical Co., St. Louis, MO) in 0.10 M NaHCO3, pH 9.0, for 60 min at 25 °C. Bacteria were washed of unbound fluorochrome with PBS by three cycles of centrifugation (13 000 rpm, 10 min). The FITC-labeled bacteria were resuspended to a concentration of 2 × 108 cells per ml in PBS, divided into equal volumes, and stored at –70 °C. Aliquots were thawed just prior to use. 2.4. Phagocytosis assay Phagocytosis was measured using a standard assay [24]. Normal human serum pooled from healthy volunteers was

J.-C. Lin et al. / Microbes and Infection 6 (2004) 1191–1198

divided into equal volumes and stored at –70 °C. Serum was thawed immediately prior to use and stored on ice until it was added to the phagocytosis assay. Briefly, for the assay, 100 µl of a neutrophil suspension (representing 1 × 106 cells), 100 µl of freshly thawed pooled normal human serum (10% v/v opsonization), and 600 µl PBS were added to sealable 10 × 75-mm Falcon™ polypropylene tubes (BD, Franklin Lakes, NJ). The suspension was pre-warmed with shaking for 5 min at 37 °C. Multiple volumes of 200 µl FITC-labeled bacteria (representing 4 × 107 colony-forming units (cfu)/ml) were added to 800 µl to produce a final volume of 1.0 ml. Each tube was capped and incubated in a shaking water bath at 37 °C, with continuous agitation for 1, 2, 5, 7.5, 10, 15, 20, 25, 30, 45, and 60 min. An unincubated tube served as 0-min tube. At each designated time, samples were removed and placed in an ice bath. The cells in each suspension were removed at 250 × g for 6 min, and the cell pellet was resuspended in 1.0 ml of ice-cold PBS and maintained at 4 °C. A 600-µl volume of the suspension was transferred into a new tube, and ethidium bromide was added to a final concentration of 50 µg/ml before measurement. Excess ethidium bromide was used to suppress the extracellular fluorescence. Bacteria that were not localized in neutrophils appeared red in color upon microscopic examination (see below). 2.5. Phagocytosis assay using flow cytometry A FACScan emitting an argon laser beam at 488 nm (Becton Dickinson Immunocytometry Systems, San Jose, CA) was used to detect FITC fluorescence. The sideways scatter (SSC) threshold was 52. The detector was set at E00, 350, and 427 for forward scatter (FSC), SSC, and fluorescence 1 (FL1-H, green), respectively. Fluorescence values were collected after gating the detector on the FSC and SSC combination. A total of 10 000 cells were processed using the Cellquest version 1.0 software (Becton Dickinson Immunocytometry Systems). Fluorescence distribution data collected using a logarithmic amplifier were displayed as single histograms for FL1-H. By processing unstained and FITC-stained bacteria phagocytosis mixtures, the boundary of positive and negative fluorescence was determined. Ingested percentage of bacteria was assessed after the addition of ethidium bromide. 2.6. Microscopic evaluation of phagocytosis Microscopic evaluation was performed as described elsewhere [24,25]. Slide preparations were made of a selection of FITC-labeled K. pneumoniae–PMN mixtures. Two hundred microliter of FITC-labeled bacteria (4 × 107 cfu/ml) were added to each tube for 1, 5, 10, 30 or 60 min. A fluorescence microscope was used to count green intracellular and red extracellular bacteria. A total of 100 cells were analyzed at each time point. A confocal microscope was used to confirm that the red bacteria were adherent and the green bacteria

1193

were ingested. Samples were prepared for confocal microscopy and examined using established methodologies. The number of bacteria that had been engulfed by neutrophils (intracellular bacteria) was determined by the number of bacteria per 100 neutrophil counts using fluorescence microscopy. The percentage of neutrophils with intracellular bacteria was determined by counting 100 consecutive neutrophils with one or more ingested bacteria. 2.7. Electron microscopic evaluation of phagocytosis Purified neutrophils were mixed with life bacteria for 30 min under the conditions described above for flow cytometry. The phagocytosis reaction was stopped by adding the reaction tube to an ice bath. Neutrophils were collected by centrifugation (250 × g, 6 min) and were washed three times with ice-cold PBS under the same centrifugation conditions. The neutrophils in the final pellet were fixed in 2.5% (v/v) glutaraldehyde (Ted Pella, CA, USA) for 60 min, washed three times with cacodylate buffer (0.1 M cacodylate with 8% sucrose, Ted Pella, CA, USA and Merck, Germany) for 10 min, post-fixed with 1% (v/v) osmium tetroxide (OsO4; Ted Pella, CA, USA) for 60 min, and washed three times (as above) with cacodylate buffer. The washed, chemically fixed cells were dehydrated in a graded ethanol series and embedded in Eponate-12 (Ted Pella, CA, USA) as described before [26]. Ultra-thin sections were stained with uranyl acetate and lead citrate (Fort Washington, PA) and examined by transmission electron microscopy (JEM 1230; JEOL, Peabody, MA) using standard operation conditions [26]. 2.8. Acapsular K. pneumoniae mutant K. pneumoniae strain DT-S (biotype edwardsii, capsular serotype K1) was kindly provided by Takeda Pharmaceuticals, Osaka, Japan [27]. DT-S was derived from K. pneumoniae DT, which was isolated from the sputum of a patient with pneumonia. Cloning of the strain was achieved by infecting mice with K. pneumoniae strain DT, then isolating the organism and repetitively infecting other mice with the harvested strain until cloning was completed. Acapsular mutant of DT-X was a non-mucoid mutant, which was isolated by subculture of strain DT-S. DT-X was confirmed to lack capsule by Indian Ink staining [27]. Both DT-S and DT-X were maintained at –80 °C in brain heart infusion (BHI) broth containing 15% glycerol. 2.9. Statistical analysis Multivariate analyses were performed to compare the differences in the adjusted PMN phagocytosis rates from 0 to 60 min between groups of interest. If significant results were found in the univariate analyses, these variables were examined using multivariate analyses. Data was expressed as least squares mean (LSMEAN) ± S.E.M. The LSMEAN represented the mean adjusted for the effect of the PMN phagocy-

1194

J.-C. Lin et al. / Microbes and Infection 6 (2004) 1191–1198 Table 2 K. pneumoniae that were selected for phagocytosis assay Serotype

Mucoid

K1 K2 Non-K1/K2 a Total

22 9 9 40

Phenotype Isolated from: Non-mucoid Liver Non-liver abscess abscess 1 14 9 1 4 6 28 2 35 30 20 50

a Serotypes included K3, K5, K6, K7, K8, K9, K15, K16, K17, K20, K21, K28, K29, K31, K32, K38, K54, K55, K57.

Fig. 1. Kinetics of phagocytosis using three individual normal human PMNs against two identical sets (K1/K2 and non-K1/K2) of K. pneumoniae isolates including K1 (23 isolates)/K2 (10 isolates) and non-K1/K2 (37 isolates). Data are mean ± standard deviations (S.D.). K1/K2 vs. non-K1/K2: P < 0.0001.

tosis rate at time 0. All statistical tests were two-sided and P-values <0.05 were considered to be statistically significant. 3. Results 3.1. Phagocytosis of K1/K2 and non-K1/K2 K. pneumoniae by human neutrophils isolated from healthy volunteers To strengthen the capsular effect, neutrophils from three healthy volunteers were isolated separately against two identical sets of K. pneumoniae with different CPS (Table 1). These included K1 (23 isolates)/K2 (10 isolates) and nonK1/K2 (37 isolates). There was a significantly higher rate of resistance to phagocytosis for K1/K2 isolates (P < 0.0001) (Fig. 1). No significant variation of the ingestion rate was found among neutrophils from the volunteers when compared to identical groups of K1/K2 or non-K1/K2 isolates. Among the 70 isolates, mucoid phenotype (M-type) was found in 40 isolates. Of these, 22 isolates were of the K1 se-

rotype, nine were of the K2 serotype, and nine belonged to non-K1/K2 serotypes. For the non-mucoid phenotype, 28 isolates were non-K1/K2 and one isolate each was K1 and K2 (Table 2). Generally, M-type isolates were more resistant to phagocytosis than non-M-type isolates (Fig. 2A). When the M-types were further differentiated with respect to their CPS, a significant difference between K1/K2 and nonK1/K2 isolates was observed. M-type K1/K2 isolates were significantly more resistant (P = 0.0029) to phagocytosis than M-type non-K1/K2 isolates (P = 0.0029), indicating that the K1/K2 CPS constituted the major factor for phagocytotic resistance (Fig. 2B). This indication was further supported by comparison of the phagocytosis rate between K1/K2 isolates with M-type and non-M-type. No significant difference in the phagocytosis rate was observed between K1/K2 isolates with M-type and non-M-type (P = 0.0924). Further comparison of phagocytosis rate showed no significant difference between M-type and non-M-type in non-K1/K2 isolates (P = 0.9019) (Fig. 2B), indicating that the effect of the mucoid phenotype on resistance to phagocytosis was increased in serotype K1/K2 isolates but not in non-K1/K2 isolates. In other words, mucoidy was not independently related to phagocytic resistance. Isolates from liver abscesses were significantly more resistant to phagocytosis than isolates from non-liver abscesses (P < 0.0001) (Fig. 3A). When further differentiated according to serotype, CPS K1/K2 isolates from liver abscess were

Fig. 2. A. Comparison of the kinetics of phagocytosis between isolates with mucoid (M-type) and non-mucoid (non-M-type) phenotype. M-type vs. non-M-type: P < 0.0001. B. Comparison of the kinetics of phagocytosis for M-type and non-M-type isolates after differentiation based on capsular serotype. M-type K1/K2 vs. M-type non-K1/K2: P = 0.0029; M-type K1/K2 vs. non-M-type K1/K2: P = 0.0924; M-type non-K1/K2 vs. non-M-type non-K1/K2: P = 0.9019.

J.-C. Lin et al. / Microbes and Infection 6 (2004) 1191–1198

1195

Fig. 3. A. Comparison of the kinetics of phagocytosis between isolates from liver abscess and non-liver abscess. Liver abscess vs. non-liver abscess: P < 0.0001. B. Comparison of the kinetics of phagocytosis for isolates from liver abscess and non-liver abscess after strengthening with their capsular serotypes. Liver abscess K1/K2 vs. liver abscess non-K1/K2: P < 0.0001; non-liver abscess K1/K2 vs. non-liver abscess non-K1/K2: P < 0.0001; liver abscess K1/K2 vs. non-liver abscess K1/K2: P = 0.1721; liver abscess non-K1/K2 vs. non-liver abscess non-K1/K2: P = 5280.

more resistant to phagocytosis than non-K1/K2 serotypes (P < 0.0001). Analysis revealed a significant difference in resistance between K1/K2 and non-K1/K2 isolates from patients without liver abscess (P = 0.0001) (Fig. 3B). These results indicated that K1/K2 isolates were generally more virulent in both liver abscess and in non-liver abscess conditions. Further comparison of non-K1/K2 isolates from patients with and without liver abscess revealed no significant difference (P = 0.5280) (Fig. 3B). These results further confirmed that the non-K1/K2 isolates were less virulent and did not specifically contribute to liver abscess. 3.2. Comparison of ingestion rates for K1/K2 and non-K1/K2 isolates using fluorescence microscopy One isolate of serotype K1 strain from liver abscesses and a non-liver-abscess control serotype K6 (ATCC700603) strain were further selected for the study of phagocytosis using methods other than flow cytometry. The fluorescence microscopic assay was repeated 10 times to minimize technical variation. Standard concentrations of bacteria (4.0 × 107 cfu) and PMNs (1 × 106) were used in each experiment. Confocal microscope analysis indicated that there was a time-dependent increase in both the number of intracellular bacteria per neutrophil and the percentage of neutrophils harboring bacteria (Fig. 4). Comparison of the number of intracellular bacteria per 100 neutrophil counts and the percentage of 100 consecutive neutrophils with one or more ingested bacteria revealed significantly higher numbers with serotype K6 bacteria than with serotype K1 bacteria (P < 0.0001) (Fig. 5A, B). The overall correlation was >0.93. 3.3. Electron microscopic evaluation of phagocytosis and intracellular killing K. pneumoniae serotype K1 was phagocytosed for 30 min. Electron microscopy failed to reveal any morpho-

logical alterations. Within the particle-containing vacuoles of the neutrophils, the cell wall structure of K1 was intact (Fig. 6), indicating resistance to lysis. In contrast, after 30 min of incubation with neutrophils, phagocytosed cells of serotype K6 displayed a less rigid cell wall, wall ruptures, and complete lysis (Fig. 6). The results were consistent with the relative increased resistance of serotype K1 isolate to phagocytosis and intracellular killing, compared to nonserotype K1 isolate. 3.4. Comparison of phagocytosis between capsulated serotype K1 isolate and acapsulated serotype K1 isogenic mutant To study the role of serotype K1 capsule in phagocytosis, serotype K1 isolates with M-type (DT-S) and isogenic acapsulated mutant with non-M-type (DT-X) were compared. Human neutrophils were isolated from six healthy volunteers and were used individually against both DT-S and DT-X. Results indicated that phagocytosis of the isogenic acapsulted mutant (DT-X) was almost 15-fold more rapid in the first 15 min than DT-S (Fig. 7). These results indicate that CPS is required for resistance to the early host defense mechanisms and contributes to the bacterial virulence in liver abscess. 4. Discussion K. pneumoniae is the most prevalent causative agent for bacterial liver abscess in many areas, and especially in Asian Pacific regions [7,9–11]. The bacterium is an emerging cause of disease in the United States [8,15,17]. A prior study conducted in Taiwan [7] documented a high prevalence of capsular serotype (CPS) K1/K2 in liver abscesses resulting from K. pneumoniae infection. Whether this observation reflected a coincidental high incidence of the serotype in the general population or whether serotype K1/K2 was actually more virulent than other serotypes remained uncertain.

1196

J.-C. Lin et al. / Microbes and Infection 6 (2004) 1191–1198

Fig. 4. The kinetics of phagocytosis of FITC fluorescence-labeled K1 and K6 K. pneumoniae based on confocal microscope analysis. PMNs and fluorescence were positioned initially and combined for analysis to confirm the ingestion of bacteria inside each cell. The number of bacteria ingested was calculated by computerized program from a confocal microscope.

The present study sought to clarify the observations of the prior investigation, by examining the phagocytosis of K. pneumoniae isolates. Since protection with non-specific immune response against bacterial infections is mediated primarily by PMNs, resistance to phagocytosis is a valid indicator of bacterial virulence. Presently, we have demonstrated that CPS serotype K1/K2 isolates are significantly more resistant to phagocytosis than non-K1/K2 isolates. This effect seems not to be a consequence of the elaboration of a copious capsule (mucoidy). The previous study had suggested that the mucoid phenotype (M-type) of K. pneumoniae was a virulence factor for infection [22]. Indeed, our initial results in the present study favored the increased phagocytic resistance of M-type isolates vs. the non-M-type isolates. However, a more detailed serological analysis failed to demonstrate a significant difference between M- and non-M-type, non-K1/K2 isolates. Furthermore, M-type K1/K2 isolates were significantly more resistant to phagocytosis than non-K1/K2. These findings suggest that mucoidy is an associated factor that was predominant in K1/K2 isolates, but which itself is not an independent influence on phagocytic resistance. The limitation of our present observation was that only two non-M-type K1 isolates were used in the comparison of M- and non-Mtype isolates. In our observation, non-mucoid K1/K2 were not frequently encountered. Further study using an increased number of non-M-type K1/K2 isolates is needed to enhance the validity of this comparison. Unfortunately, this limitation

was due to the fact that non-M K1/K2 isolates are rare in the general population. In a previous collection, only two among 158 serotype K1/K2 isolates were non-M K1/K2. Perhaps molecular cloning or knockout of specific gene for M-type may help to delineate the role of this factor. Isolates from liver abscess were significantly more resistant to phagocytosis than non-abscess isolates. This phenomenon was caused mainly by the high prevalence of K1/K2 serotypes in isolates from liver abscess patients and the lower prevalence of the K1/K2 serotypes in isolates from hospitalacquired infection [4]. Taken together, the present results support the contention that CPS (serotype K1/K2) contributes to virulence but not as the sole cause of liver abscess. According to our data, a significant difference in phagocytosis and intracellular killing was still observed between K1/K2 and non-K1/K2 isolates from patients with liver abscess. If specific CPS was the sole factor involved in the development of liver abscess, all strains with different CPS isolated from liver abscess would exhibit no significant differences in phagocytotic properties. Our findings thus imply that bacterial virulence factors other than CPS, or factors other than bacteria, are involved in the development of liver abscess. A previous study implicated adhesins, whose activity can be inhibited by mannose as a virulence factor [28]. Whether non-K1/K2 isolates from liver abscess contain similar adhesins remains to be assessed. The resistance of CPS K1 to neutrophil-mediated phagocytosis was independently confirmed using microscopic and

J.-C. Lin et al. / Microbes and Infection 6 (2004) 1191–1198

1197

Fig. 7. Comparison of kinetics of phagocytosis between serotype K1 isolate and its isogenic acapsulated mutant. Data are mean ± S.D. of six independent experiments.

Fig. 5. A. Comparison of phagocytosis between K1 and K6 isolates according to the number of intracellular bacteria in neutrophils. Phagocytosis was determined by number of bacteria per 100 neutrophil counts using a fluorescence microscope. Data are mean ± S.D. of 10 independent experiments. Multivariate analyses, P < 0.0001. B. Comparison of phagocytosis between K1 and K6 isolates was based on the percentage of neutrophils with intracellular bacteria, determined by counting 100 consecutive neutrophils with ≥1 ingested bacterium. Data are mean ± S.D. of 10 independent experiments. Multivariate analyses, P < 0.0001.

phils, while the control K6 isolate was susceptible to phagocytosis. Whether the intracellular proliferation of K1 isolates occurs during phagocytosis or intracellular killing requires further study. However, the present findings imply that the resistance to neutrophil phagocytosis and intracellular killing, resulting in survival in the circulatory system, could lead to metastasis to the eye. Since serotype K1/K2 contributes to the impaired PMN-mediated blood clearance of bacteria in liver abscess patients, this finding provides a possible explanation of why only K1/K2 isolates were found in cases complicated by endophthalmitis [7]. In summary, the observations from the present experimental model and from our previously published epidemiological surveys [4,7,21] strongly suggest that K. pneumoniae CPS serotype K1 or K2 expresses a virulence factor that contributes to the development of liver abscess as well as complication by endophthalmitis. However, the reason for the high prevalence of this specific disease in DM and in the Asia– Pacific region has yet to be determined. Our data identifying these two factors provide an important basis to further delineate the mechanisms responsible for K. pneumoniae liver abscess and endophthalmitis complication. Further genetic/genomic study on specific CPS-correlated genes is important to complete our understanding of the mechanisms of virulence as well as the causes of liver abscess.

Acknowledgements

Fig. 6. Comparison of the viable bacterial K1 and K6 isolates after 30 min of phagocytosis and intracellular killing examined by electron microscopy.

flow cytometry. An intriguing finding in the electron microscopic examination was the resistance to intracellular killing of viable K1 bacteria after 30 min in the presence of neutro-

We thank Professor Tetusya Matsumoto and Keizo Yamaguchi, Department of Microbiology, Toho University, Tokyo, Japan for kindly supplying the K1 mutant isolate. This work was supported by grants from National Health Research Institutes and National Science Council (NSC 912314-B-016-047 & NSC 92-2314-B-016-026), Taiwan. Part of this work was presented in ICAAC, 43rd ASM’s annual meeting on infectious diseases, Chicago, IL, USA, B-1053, 2003. Abs. no. B1053.

1198

J.-C. Lin et al. / Microbes and Infection 6 (2004) 1191–1198

References [1]

[2]

[3]

[4]

[5]

[6]

[7]

[8] [9]

[10] [11]

[12]

[13]

D.L. Paterson, W.C. Ko, A. Von Gottberg, J.M. Casellas, L. Mulazimoglu, K.P. Klugman, R.A. Bonomo, L.B. Rice, J.G. McCormack, V.L. Yu, Outcome of cephalosporin treatment for serious infections due to apparently susceptible organisms producing extendedspectrum beta-lactamases: implications for the clinical microbiology laboratory, J. Clin. Microbiol. 39 (2001) 2206–2212. L.K. Siu, P.L. Lu, P.R. Hsueh, F.M. Lin, S.C. Chang, K.T. Luh, et al., Bacteremia due to extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae in a pediatric oncology ward: clinical features and identification of different plasmids carrying both SHV-5 and TEM-1 genes, J. Clin. Microbiol. 37 (1999) 4020–4027. W.R. Jarvis, V.P. Munn, A.K. Highsmith, D.H. Culver, J.M. Hughes, The epidemiology of nosocomial infections caused by Klebsiella pneumoniae, Infect. Control 6 (1985) 68–74. R.W. Tsay, L.K. Siu, C.P. Fung, F.Y. Chang, Characteristics of bacteremia between community-acquired and nosocomial Klebsiella pneumoniae infection: risk factor for mortality and the impact of capsular serotypes as a herald for community-acquired infection, Arch. Intern. Med. 162 (2002) 1021–1027. J.A. Korvick, C.S. Bryan, B. Farber, T.R. Beam Jr., L. Schenfeld, R.R. Muder, et al., Prospective observational study of Klebsiella bacteremia in 230 patients: outcome for antibiotic combinations versus monotherapy, Antimicrob. Agents Chemother. 36 (1992) 2639– 2644. E.N. Barton, H. Daisley, D.T. Gilbert, L. Roberts, Diabetes mellitus and Klebsiella pneumoniae liver abscess in adults, Trop. Geogr. Med. 43 (1991) 100–104. C.P. Fung, F.Y. Chang, S.C. Lee, B.S. Hu, B.I. Kuo, C.Y. Liu, M. Ho, L.K. Siu, A global emerging disease of Klebsiella pneumoniae liver abscess: is serotype K1 an important factor for complicated endophthalmitis? Gut 50 (2002) 420–424. N. Hansen, T. Vargish, Pyogenic hepatic abscess: a case for open drainage, Am. Surg. 59 (1993) 219–222. H. Okano, K. Shiraki, H. Inoue, T. Kawakita, N. Yamamoto, M. Deguchi, K. Sugimoto, T. Sakai, S. Ohmori, K. Murata, T. Nakano, Clinicopathological analysis of liver abscess in Japan, Int. J. Mol. Med. 10 (2002) 627–630. K.G. Yeoh, I. Yap, S.T. Wong, A. Wee, R. Guan, J.Y. Kang, Tropical liver abscess, Postgrad. Med. J. 73 (1997) 89–92. V. Wiwanitkit, N. Suwansaksri, J. Suwansaksri, Causative agents of liver abscess in those with liver cirrhosis: a 10-year case review of hospitalized patients in Thailand, Ann. Trop. Med. Parasitol. 96 (2002) 513–516. S.T. Lindstrom, P.R. Healey, S.C. Chen, Metastatic septic endophthalmitis complicating pyogenic liver abscess caused by Klebsiella pneumoniae, Aust. N. Z. J. Med. 27 (1997) 77–78 [letter]. S. Ohmori, K. Shiraki, K. Ito, H. Inoue, T. Ito, T. Sakai, K. Takase, T. Nakano, Septic endophthalmitis and meningitis associated with Klebsiella pneumoniae liver abscess, Hepatol. Res. 22 (2002) 307– 312.

[14] T.Y. Wong, S.I. Chiu, M.K. So, M.K. Tsang, S.T. Lai, K.K. Tse, I.Y. Io, Septic metastatic endophthalmitis complicating Klebsiella pneumoniae liver abscess in a non-diabetic Chinese man, Hong Kong Med. J. 7 (2001) 303–306. [15] S.R. Ayinala, M. Vulpe, M. Azaz, H. Cohen, S.S. Donelson, M. Lee, Pyogenic liver abscesses due to Klebsiella pneumoniae in a diabetic patient, J. Miss. State Med. Assoc. 42 (2001) 67–70. [16] M.F. Cobo, M.L. Aliaga, M.F. Diaz, J.D. Mediavilla Garcia, J.P. Arrebola Nacle, F.M. De la Rosa, Liver abscess caused by Klebsiella pneumoniae in diabetic patients, Rev. Clin. Esp. 199 (1999) 517–519. [17] M. Saccente, Klebsiella pneumoniae liver abscess, endophthalmitis, and meningitis in a man with newly recognized diabetes mellitus, Clin. Infect. Dis. 29 (1999) 1570–1571. [18] H.R. Liao, H.W. Lee, H.S. Leu, B.J. Lin, C.J. Juang, Endogenous Klebsiella pneumoniae endophthalmitis in diabetic patients, Can. J. Ophthalmol. 27 (1992) 143–147. [19] S.S. Yarng, C.L. Hsieh, T.L. Chen, Vitrectomy for endogenous Klebsiella pneumoniae endophthalmitis with massive subretinal abscess, Ophthalmic Surg. Lasers 28 (1997) 147–150. [20] D.M. Bamberger, B.L. Herndon, J. Fitch, A. Florkowski, V. Parkhurst, Effects of neutrophils on cefazolin activity and penicillin-binding proteins in Staphylococcus aureus abscesses, Antimicrob. Agents Chemother. 46 (2002) 2878–2884. [21] C.P. Fung, B.S. Hu, F.Y. Chang, S.C. Lee, B.I. Kuo, M. Ho, et al., A 5-year study of the seroepidemiology of Klebsiella pneumoniae: high prevalence of capsular serotype K1 in Taiwan and implication for vaccine efficacy, J. Infect. Dis. 181 (2000) 2075–2079. [22] X. Nassif, J.M. Fournier, J. Arondel, P.J. Sansonetti, Mucoid phenotype of Klebsiella pneumoniae is a plasmid-encoded virulence factor, Infect. Immun. 57 (1989) 546–552. [23] P.A. Campbell, B.P. Canono, D.A. Dervets, Isolation and functional analysis of neutrophils, in: Current Protocals in Immunology, Wiley, New York, 1996. [24] M. Heinzelmann, S.A. Gardner, M. Mercer-Jones, A.J. Roll, H.C. Polk Jr., Quantification of phagocytosis in human neutrophils by flow cytometry, Microbiol. Immunol. 43 (1999) 505–512. [25] D.A. Drevets, P.A. Campbell, Macrophage phagocytosis: use of fluorescence microscopy to distinguish between extracellular and intracellular bacteria, J. Immunol. Methods 142 (1991) 31–38. [26] J.J. Wang, Y. Lu, L. Ratner, Particle assembly and Vpr expression in human immunodeficiency virus type 1-infected cells demonstrated by immunoelectron microscopy, J. Gen. Virol. 75 (Pt 10) (1994) 2607– 2614. [27] K. Yoshida, T. Matsumoto, K. Tateda, K. Uchida, S. Tsujimoto, K. Yamaguchi, Role of bacterial capsule in local and systemic inflammatory responses of mice during pulmonary infection with Klebsiella pneumoniae, J. Med. Microbiol. 49 (2000) 1003–1010. [28] C. Pruzzo, C.A. Guzman, L. Calegari, G. Satta, Impairment of phagocytosis by the Klebsiella pneumoniae mannose-inhibitable adhesinT7 receptor, Infect. Immun. 57 (1989) 975–982.