Entry of Sanfetrinem into Human Polymorphonuclear Granulocytes and Its Cell-Associated Activity against Intracellular, Penicillin-Resistant Streptococcus pneumoniae

Abstract

The entry of antibiotics into phagocytes is necessary for activity against intracellular pathogens. The ability of sanfetrinem, the first member of a new class of antibiotics, to penetrate human polymorphonuclear granulocytes and its consequences upon subsequent phagocytosis and killing of ingested penicillin-resistant Streptococcus pneumoniae have been evaluated. Sanfetrinem penetrated into human polymorphonuclear leukocytes (PMNs) at all concentrations tested, with cellular concentration/extracellular concentration ratios of 6.6 to 5.03 and 4.21 when sanfetrinem was used at 0.25 to 0.5 and 1 μg/ml, respectively, within 30 min of incubation. The uptake was complete within 5 min and was not energy dependent, since it was not affected by cell viability, environmental temperature, or the addition of a metabolic inhibitor. At a concentration of one-half the MIC, sanfetrinem significantly enhanced human PMN phagocytosis and increased intracellular bactericidal activity against penicillin-resistant S. pneumoniae. Following preexposure of PMNs to a concentration of one-half the MIC of sanfetrinem, there was a significant increase in both phagocytosis and killing compared with that for the controls, indicating the ability of sanfetrinem to interact with biological membranes and remain active within PMNs. Preexposure of streptococci to sanfetrinem made penicillin-resistant S. pneumoniae more susceptible to the bactericidal mechanisms of human PMNs than untreated organisms.

It is only in recent years that the emerging concept of immunomodulation by antimicrobial agents has received worldwide interest (4, 11, 15, 17, 24, 30, 34). Thus, the current trend of therapy requires the use of antibiotics which combine a high level of in vitro antibacterial activity with the capacity to act in concert with the immune system in a way that potentiates the host’s defense mechanisms. Among the several multifacetted aspects of the antibiotic-phagocyte interaction, only antibiotic entry into phagocytes and, subsequently, its bioactivity are considered to be clinically relevant and beneficial for the treatment of infections caused by pathogens that are capable of survival and replication within phagocytic cells, constituting a significant cause of human infections. Several antimicrobial agents have been reported to accumulate to high levels in phagocytes: macrolides, fluoroquinolones, and some antitubercular drugs (1, 5, 8, 10, 14, 25, 35, 37). Unfortunately, among the cell wall-acting antibiotics, most β-lactams do not efficiently penetrate phagocytes and the newer β-lactam drugs, such as cefotaxime, ceftizoxime, cefonicid, ceftriaxone, and ceftazidime, are similar to penicillin in their relative inabilities to penetrate phagocytes (4, 11, 22, 23). Among the cell wall inhibitors, only teicoplanin (3) and carbapenem antibiotics, such as imipenem and meropenem (6, 12, 13, 21), have been shown to have high levels of penetration into different phagocytic cells. Although imipenem binds to phagocytes, the concentration of cell-associated drug declines steadily during an incubation period of 1 h (21); on the other hand, meropenem is able to penetrate human phagocytes and remain intracellularly active (13).

In the present study we have investigated the uptake of sanfetrinem, the first member of a new class of antibiotics (trinems), by human polymorphonuclear cells (PMNs), the crucial phagocytes that offer protection against most bacterial pathogens, and further examined the in vitro consequences of its uptake on subsequent PMN activities toward a penicillin-resistant strain of Streptococcus pneumoniae, a pathogen that is known to be a leading cause of invasive infections and that is often associated with considerable morbidity and mortality (18, 28).

(This work was presented in part at the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Ontario, Canada, 28 September to 1 October 1997 [14a].)

MATERIALS AND METHODS

Bacteria.

The S. pneumoniae strain used in this study was strain 4636, serotype 19, derived from a human infection; it is virulent in mice and is resistant to penicillin (MICs of penicillin G and amoxicillin, 4 μg/ml for 10⁵ CFU/ml).

Antibiotic.

Sanfetrinem was kindly provided by Glaxo Wellcome, Verona, Italy. The antibiotic solutions were freshly prepared for each experiment. Antibiotic susceptibility testing was performed by the standard dilution method in Mueller-Hinton broth (Unipath, Milan, Italy).

PMNs.

Blood was drawn from healthy subjects who gave their informed consent. Peripheral venous blood was collected in sterile evacuated blood collection tubes containing lithium heparin (15 U/ml of blood), and the contents were allowed to settle by gravity at room temperature for 30 min in 2.5% dextran (molecular weight, 70,000; Pharmacia S.p.A., Milan, Italy) in normal saline (1:1 ratio). The leukocyte-rich plasma supernatant was carefully layered on Ficoll-Paque (Pharmacia) and was then centrifuged twice at 1,200 × g for 15 min; to obtain pure PMNs, residual erythrocytes were lysed by hypotonic shock for 30 s in sterile distilled water, and then the PMNs were further centrifuged. After being counted in a Bürker cell counting chamber, the PMN density was adjusted to 10⁶ cells/ml in phosphate-buffered saline supplied with 1% glucose and 0.1% human albumin (Sigma); the PMNs were placed in sterile plastic tubes, treated with RPMI 1640 (Gibco Laboratories, Grand Island, N.Y.), supplemented with 10% fetal calf serum (Gibco), and incubated for various periods of time at 37°C in a shaking water bath (150 rpm). The viability was assayed by trypan blue exclusion and was greater than 95%; a viability test was done before and after each experiment. The time between collection of blood and the beginning of the experiments did not exceed 3 h; the interval between PMN harvest and the start of the experiments was less than 30 min.

Technique and determination of sanfetrinem uptake by PMNs.

¹⁴C-labelled sanfetrinem (specific activity, 524 MBq/mmol) was added to phagocytes at concentrations of 0.25, 0.5, 1, 2, 4, and 8 μg/ml. The PMN-antibiotic mixtures were then incubated at 37°C in a shaking water bath. At intervals the supernatant was removed by centrifugation at 1,200 × g for 5 min; the pellet was then resuspended in phosphate saline, and the mixture was centrifuged at 1,200 × g for 5 min to remove extracellular antibiotic. The radioactivity both in the supernatant and in the wash buffer was measured. The concentration of unbound intracellular drug was determined by lysing the PMNs in 1 ml of distilled water. Aliquots of extracellular antibiotic and intracellular unbound sanfetrinem were placed in a scintillation liquid (Atomlight; Packard, Milan, Italy) and counted by liquid scintillation spectrophotometry.

To determine the amount of intracellular bound drug, 0.5-ml aliquots of each lysate were diluted with 4.5 ml of 1 M KH₂PO₄, and the pH was adjusted to 4 with 1 N HCl (3, 13). The diluted sample was extracted with 15 ml of ethyl acetate for 10 min at 300 inversions per min on a Continental Alter 2644 shaker. After centrifugation at 1,800 × g for 10 min, 14 ml of the organic phase was evaporated to dryness under a stream of nitrogen. Aliquots of intracellular bound sanfetrinem were placed in a scintillation liquid (Atomlight) and counted by liquid scintillation spectrophotometry. Measurements of uptake by the cells were done in triplicate at each interval.

The intracellular concentration of antibiotic/extracellular concentration of antibiotic ratio (the C/E distribution ratio) was calculated from the following relationship (13): C/E ratio = (counts per minute/10⁶ cells × 200)/(counts per minute/ml of medium).

The total intracellular sanfetrinem concentration (C) used in the calculation of the C/E ratio was the sum of the bound and unbound fractions of each drug. The factor 200 was obtained by dividing 1 ml by 0.005 ml (volume of 10⁶ cells, assuming them to be spherical with a radius of 10⁻³ cm) so that the values could be expressed in terms of counts per minute/milliliter of fluid in both the numerator and the denominator. The distribution ratio is a unitless value. The counts per minute of each test sample was always compared with the counts per minute of a standard solution of the same drug. The degree of penetration was also expressed in micrograms per milliliter by using the following binomial (13): a = x + y and b = x/y, where x and y were the intracellular and extracellular drug concentrations to be determined, respectively; a was the known antibiotic concentration tested; and b was the mean value of the C/E ratio.

Characterization of sanfetrinem uptake.

To elucidate the mechanism of sanfetrinem uptake by human PMNs, we examined the uptake by PMNs killed by exposure to 10% (vol/vol) formalin for 30 min. These cells were then washed twice by centrifugation at 1,200 × g for 5 min and were suspended in fresh medium. Sanfetrinem uptake was also determined at 4 and 37°C. Furthermore, the cells in RPMI 1640 were incubated with and without the metabolic inhibitor ouabain (Sigma) at 10⁻³ M for 30 min at 37°C in a shaking water bath before antibiotic concentration determinations.

Opsonization procedure.

Serum from a pool of healthy volunteers was used. After the blood had been allowed to clot for 1 h at room temperature, the serum was collected by centrifugation for 20 min at 1,100 × g, aliquoted, and stored at −70°C until use. Human pooled serum was used unheated (intact complement system). Streptococci (5 × 10⁷ CFU/ml) were incubated for different incubation times (15, 30, or 60 min) at 37°C with 10% human pooled serum; opsonization was stopped by the addition of 2.5 ml of ice-cold phosphate saline, and serum was removed by centrifugation at 2,000 × g for 10 min; the bacteria were then resuspended in fresh medium to a final concentration of 2 × 10⁷ CFU/ml, as confirmed by obtaining colony counts in triplicate.

Radioactive labelling protocol.

S. pneumoniae was grown from overnight chocolate agar colonies placed in 10 ml of Todd-Hewitt (TH) broth (Unipath) containing 150 μCi of ³H-uracil (specific activity, 1,165.5 GBq/mmol; Du Pont de Nemours, NEN Products, Milan, Italy) in a 5% CO₂ incubator for 4 h. The radiolabelled bacteria were centrifuged several times with TH broth and were resuspended in fresh medium and adjusted to an optical density at 620 nm of 0.15 to yield 2 × 10⁷ CFU/ml, as confirmed by obtaining colony counts in triplicate.

Phagocytosis assay.

In all the experiments, the bacterium:PMN ratio was 10:1. Aliquots of 1.0 ml of streptococci (2 × 10⁷ CFU) in RPMI 1640 with 10% fetal calf serum were added to PMNs in sterile plastic tubes (10⁶ cells), and the tubes were then incubated at 37°C in a shaking water bath. After incubation for periods of 30, 60, or 90 min the tubes were centrifuged at 1,200 × g for 5 min; the pellet was then resuspended in phosphate saline, and the mixture was centrifuged at 1,200 × g for 5 min to remove the free streptococci. The cells were then resuspended in 1 ml of sterile distilled water for 5 min, and 100-μl samples of this suspension were placed in scintillation fluid (Atomlight) and counted by liquid scintillation spectrophotometry. Radioactivity was expressed as the counts per minute/sample. The percentage of phagocytosis at a given sampling time (14) was calculated as follows: percent phagocytosis = [(counts per minute in PMN pellet)/(counts per minute in total bacterial pellet)] × 100.

Measurement of antimicrobial activity of PMNs.

In all the experiments, the bacterium:PMN ratio was 10:1. Aliquots of 1.0 ml of streptococci (2 × 10⁷ CFU) and PMNs in sterile plastic tubes (10⁶ cells) were incubated in RPMI 1640 at 37°C in a shaking water bath for 30 min to allow phagocytosis to proceed. The PMN-bacterium mixtures were centrifuged at 1,200 × g for 5 min and washed with phosphate saline to remove the free extracellular bacteria. An aliquot of the cells containing streptococci was taken, the cells were lysed by adding sterile water, and a viable count of intracellular bacteria was performed (time zero). The cells were then incubated further, and at intervals (time x) the viable counts of the surviving intracellular bacteria were measured in the same way. The PMN killing values were expressed as the survival index (SI), which was calculated by adding the number of surviving bacteria at time zero to the number of survivors at time x and dividing by the number of survivors at time zero (14). According to this formula, if bacterial killing was 100% effective, the SI would be 1.

Effect of sanfetrinem on PMN functions.

The effects of sanfetrinem on the phagocytosis and intracellular killing of S. pneumoniae by PMNs were investigated by incubating the bacteria and the phagocytes at 37°C in a shaking water bath for periods of 30, 60, or 90 min in the presence of one-half the MIC of sanfetrinem; antibiotic-free controls were also included. Phagocytosis and intracellular killing were assessed by the methods described above. The distinction between any effect of sanfetrinem on the bacteria and the PMNs was made by the preexposure of each of them to the antibiotic before they were incubated together. The bacteria were inoculated in 10 ml of TH broth (control) or TH broth containing one-half the MIC of sanfetrinem. After incubation for 1 h at 37°C, the suspension was centrifuged at 1,600 × g for 15 min, and the bacteria were washed in phosphate saline to remove the drug and were then adjusted to 2 × 10⁷ CFU/ml. Similarly, either 1 ml of a solution containing sanfetrinem or 1 ml of phosphate saline (control) was added to the PMNs in sterile plastic tubes, and the tubes were then incubated at 37°C in a shaking water bath. After 1 h the cells were centrifuged at 1,200 × g for 5 min, and the pellet was suspended in phosphate saline and centrifuged at 1,200 × g for a further 5 min to remove the antibiotic. Preexposed streptococci were added to PMNs and the streptococci were added to preexposed PMNs, and the tubes were incubated at 37°C in a shaking water bath for periods of 30, 60, or 90 min. The phagocytic and bactericidal activities of the PMNs were determined by the methods described above. A control system was assayed in parallel with assays of the variables described above. Every test was carried out in triplicate, and the results were compared with those obtained with control systems which contained no sanfetrinem. Results are expressed as the means and standard errors of the means (SEMs) for five separate experiments.

Statistical analysis.

Statistical evaluation of the differences between test and control results were performed by an analysis of variance by Tukey’s test.

RESULTS

Sanfetrinem uptake by human PMNs.

In all the experiments the viability of PMNs remained unchanged throughout. The mean ± SEM C/E ratios for sanfetrinem with PMNs are presented in Table 1, indicating that when sanfetrinem is incubated at various concentrations with human PMNs at 37°C, the drug markedly penetrated the phagocytes. The penetration was almost complete at the lower concentrations tested. In fact, with an extracellular concentration of 0.25 μg/ml, the C/E ratios were greater than 5 throughout the entire period of incubation; the intracellular concentration was relatively stable over time (0.21 μg/ml) and was close to the exposure concentration (0.25 μg/ml). At both 0.5 and 1 μg/ml (1 μg/ml is the peak level in serum) the intracellular concentrations were also quite near the exposure concentration for the entire period of observation, with values of 0.37 to 0.4 and 0.7 to 0.8 μg/ml, respectively (Table 1). Increasing the exposure concentration resulted in a progressively lower percentage of the antibiotic entering the intracellular compartment, although in absolute terms the concentrations increased. Thus, at 2, 4, and 8 μg/ml, sanfetrinem achieved, within 30 min, intra-PMN concentrations of 1.5, 3.06, and 5.9 μg/ml, respectively (data not shown). Sanfetrinem bound rapidly (within 5 min) to human PMNs; interestingly, the larger amount of cell-associated drug was registered at 30 min for all drug concentrations tested (Table 1). An analysis of the free drug/bound drug ratios showed that many binding sites were available for sanfetrinem, and the ratio of free sanfetrinem/bound sanfetrinem increased with increasing concentration: 0.6 at 0.25 μg/ml, 0.8 at 0.5 μg/ml, and 0.9 at 1 μg/ml.

TABLE 1.

Uptake of ¹⁴C-sanfetrinem by human polymorphonuclear leukocytes after exposure of cell cultures to various concentrations of the antibiotic and for various time periods

Time (min)	Uptake of 14C-sanfetrinem when it was in the medium at the following concn at time zeroa:
	0.25 μg/ml		0.5 μg/ml		1 μg/ml
	C/E ratio	IC (μg/ml)	C/E ratio	IC (μg/ml)	C/E ratio	IC (μg/ml)
5	5.28 ± 0.42	0.210	2.81 ± 0.71	0.369	2.71 ± 0.4	0.731
10	4.22 ± 0.80	0.203	2.29 ± 0.04	0.349	2.14 ± 0.06	0.682
20	5.30 ± 0.40	0.210	3.42 ± 0.18	0.387	3.09 ± 0.34	0.756
30	6.6 ± 0.23	0.217	5.03 ± 0.2	0.42	4.21 ± 0.38	0.809
60	5.53 ± 0.15	0.211	2.82 ± 0.04	0.37	3.25 ± 0.12	0.765

^aAll data are means ± SEMs. IC, intracellular concentrations. 

Characterization of sanfetrinem uptake.

Cellular uptake of sanfetrinem was independent of cell viability, physiological environmental temperature, or metabolic requirement (Table 2). In fact, the entry of sanfetrinem was not different when formalin-killed cells at 37°C or viable human PMNs at 4°C were used. The presence of ouabain, an inhibitor of cell membrane sodium or potassium transporting ATPase system, did not affect the intracellular entry of this antibiotic. Differences among groups of PMNs were not statistically significant.

TABLE 2.

Influences of cell viability, metabolic inhibitor, and environmental temperature on the C/E ratio for sanfetrinem

Sanfetrinem concn (μg/ml)	Time (min)	C/E ratio
		Controlsa	Formalin-killed cells at 37°C	Ouabain-treated cells at 37°C	Viable cells at 4°C
0.25	10	4.22 ± 0.80	4.12 ± 0.09	4.26 ± 0.14	4.2 ± 0.13
	30	6.6 ± 0.23	6.4 ± 0.55	6.81 ± 0.90	6.18 ± 0.05
	60	5.53 ± 0.15	6.13 ± 0.20	5.9 ± 0.35	5.6 ± 0.17
0.5	10	2.29 ± 0.04	2.37 ± 0.28	2.31 ± 0.09	2.75 ± 0.22
	30	5.03 ± 0.2	4.95 ± 0.37	5.05 ± 0.13	4.88 ± 0.36
	60	2.82 ± 0.04	2.73 ± 0.26	2.8 ± 0.37	2.96 ± 0.27
1	10	2.14 ± 0.06	2.27 ± 0.20	2.43 ± 0.11	2.34 ± 0.19
	30	4.21 ± 0.38	4.7 ± 0.15	4.19 ± 0.03	4.4 ± 0.12
	60	3.25 ± 0.12	2.59 ± 0.23	3.14 ± 0.29	3.38 ± 0.24

^aViable human PMNs at 37°C. 

Effect of sanfetrinem on human PMN functions.

The MIC of sanfetrinem for the penicillin-resistant strain of S. pneumoniae was found to be 0.5 μg/ml, equivalent to its minimum bactericidal concentration. The presence of one-half the MIC of sanfetrinem affected the rate of human granulocyte phagocytosis, resulting in an increased percentage of ingested streptococci over all three time points in comparison with that for the control (P < 0.05; Table 3). The incubation of PMNs containing intracellular S. pneumoniae in an antibiotic-free control culture resulted in an increase in viable counts: SIs were greater than 2, which is compatible with intracellular survival (Table 3). The sanfetrinem added to the PMNs after the phagocytosis had occurred significantly enhanced the phagocytes’ intracellular microbicidal activity against ingested penicillin-resistant streptococci: during the 90-min period the intracellular bacterial load was reduced by 70% (P < 0.01; Table 3). The in vitro susceptibilities of the extracellular bacteria and the viable organisms recovered from lysed PMNs were indistinguishable, indicating that the surviving streptococci had not acquired resistance to sanfetrinem following exposure to the antibiotic (data not shown).

TABLE 3.

Effect of one-half the MIC of sanfetrinem on human PMN phagocytosis and intracellular killing of S. pneumoniae^a

Time (min)	% Phagocytosis		SI
	Controls	Sanfetrinem	Controls	Sanfetrinem
30	12.51 ± 1.24	18.23b ± 1.1	>2	1.30c ± 0.34 (70)d
60	10.41 ± 1.35	16.09b ± 1.01	>2	1.36c ± 0.2 (64)
90	8.97 ± 0.99	10.45b ± 1.2	>2	1.30c ± 0.19 (70)

^aValues are means ± SEMs. 

^bSignificantly different (P < 0.05) from the controls. 

^cSignificantly different (P < 0.01) from the controls. 

^dValues in parentheses are the percentages of the initial bacterial population killed by PMNs in the presence of the antibiotic. 

To investigate the direct effect of sanfetrinem on the phagocyte functions, human PMNs were preincubated for 1 h with one-half the MIC of the drug. After withdrawal of the antibiotic, bacterial uptake and microbicidal activity were determined. As documented by the results in Table 4, both the phagocytosis and the intracellular killing of penicillin-resistant S. pneumoniae were influenced by the direct effect of sanfetrinem on the human PMNs compared with the results for the controls. In fact, the streptococci were phagocytosed by drug-preexposed PMNs at a rate that was twice the rate of phagocytosis by unexposed cells, and a significant stimulation of granulocyte killing of intracellular phagocytosed bacteria occurred, in contrast to the killing observed with untreated phagocytes (P < 0.01; Table 4).

TABLE 4.

Effect of 1 h of preexposure of human PMNs to one-half the MIC of sanfetrinem on phagocytosis and intracellular killing of S. pneumoniae^a

Time (min)	% Phagocytosis		SI
	Controls	Sanfetrinem	Controls	Sanfetrinem
30	12.51 ± 1.24	16.71b ± 0.96	>2	1.66c ± 0.26 (44)d
60	10.41 ± 1.35	16.81b ± 1.5	>2	1.50c ± 0.2 (50)
90	8.97 ± 0.99	17.02c ± 1.56	>2	1.24c ± 0.23 (76)

^aValues are means ± SEMs. 

^bSignificantly different (P < 0.05) from the controls. 

^cSignificantly different (P < 0.01) from the controls. 

^dValues in parentheses are the percentages of the initial bacterial population killed by PMNs in the presence of the antibiotic. 

Preexposure of penicillin-resistant S. pneumoniae to one-half the MIC of sanfetrinem for 1 h during the growth phase had no effect upon the phagocytic capacity of human PMNs for the entire period of observation: streptococci treated with sanfetrinem were phagocytosed within 90 min of incubation at the same percentage as that registered for untreated S. pneumoniae (Table 5). While phagocytosed, sanfetrinem-untreated streptococci (controls) remained viable in PMNs, the number of intracellular sanfetrinem-exposed bacteria decreased with time: at 30, 60, and 90 min the SIs were 1.36, 1.37, and 1.27, respectively (P < 0.01; Table 5), indicating a phagocytic microbicidal activity greater than that of the control system. In the absence of PMNs there was no significant difference between the growth of streptococci which had previously been exposed to the sub-MIC of sanfetrinem and the growth of bacteria which had not undergone preexposure (data not shown).

TABLE 5.

Effect of 1 h of preexposure of S. pneumoniae to one-half the MIC of sanfetrinem on human PMN phagocytosis and intracellular killing^a

Time (min)	% Phagocytosis		SI
	Controls	Sanfetrinem	Controls	Sanfetrinem
30	12.51 ± 1.24	14.01 ± 1.32	>2	1.36b ± 0.14 (64)c
60	10.41 ± 1.35	11.42 ± 0.5	>2	1.37b ± 0.31 (63)
90	8.97 ± 0.99	8.62 ± 1.35	>2	1.27b ± 0.21 (73)

^aValues are means ± SEMs. 

^bSignificantly different (P < 0.01) from the controls. 

^cValues in parentheses are the percentages of the initial bacterial population killed by PMNs in the presence of the antibiotic. 

Opsonization.

Serum preopsonization of the streptococci had no effect on either phagocytosis or intracellular killing by PMNs in comparison with the effect observed with nonopsonized microorganisms (data not shown). To find out a possible difference in uptake and killing, the experiments were also performed with longer opsonization times (30 to 60 min), but we failed to show such a difference. The addition of sanfetrinem to opsonized S. pneumoniae in the presence of PMNs resulted in a picture similar to that seen with nonopsonized bacteria (data not shown).

DISCUSSION

Sanfetrinem, the first member of a new class of tricyclic β-lactam antibiotics (trinems), is a potent agent with a broad antibacterial spectrum that encompasses a wide range of gram-negative, gram-positive, and anaerobic pathogenic bacteria (16, 32). Against gram-negative bacteria, sanfetrinem shows an activity broadly comparable to that of the cephalosporins, but it has superior potency against a range of gram-positive organisms, including methicillin-resistant Staphylococcus aureus and multidrug-resistant enterococci (32). It is completely stable in the presence of all clinically relevant β-lactamases and exhibits greater antimicrobial activity or a wider spectrum of activity than those of the semisynthetic penicillins even when its activity is potentiated by the addition of a β-lactamase inhibitor (16, 32). Unlike imipenem, sanfetrinem is not degraded by human renal dehydropeptidases (29, 32).

From the results of this study it emerges that sanfetrinem combines all of the properties mentioned above with the capability of penetrating the phagocytes in the drug’s microbiologically active form, directly enhancing phagocytic activities, and acting on the intracellular, replicating penicillin-resistant S. pneumoniae. The level of penetration of some antimicrobial agents is limited, while some others are taken up well (5, 17, 19–23, 37). The β-lactam group is generally considered to penetrate the phagocyte membrane poorly. In fact, most β-lactam drugs, the newer ones included, have C/E ratios of less than 1, and it has generally been accepted that they have little effect on intracellular bacteria. Among cell wall-acting antibiotics, only carbapenems have been shown to penetrate the phagocytes. Imipenem binds rapidly to PMNs, but the amount of cell-associated drug progressively declines within 1 h; this is probably related to a rapid binding to the cell membrane followed by dissociation, extracellular hydrolysis of the antibiotic, or cellular metabolism of the drug (21). The uptake of meropenem was shown to be through passive mechanisms, and phagocyte-associated drug reduced the numbers of viable intracellular staphylococci (13). The experiments described here indicate that sanfetrinem efficiently penetrated human PMNs at all concentrations tested, with C/E ratios of 6.6, 5.03, and 4.21 for sanfetrinem at 0.25, 0.5, and 1 μg/ml, respectively, within 30 min of incubation (Table 1). The penetration, expressed as the intracellular concentration (in micrograms per milliliter), was almost complete at the lower concentrations tested (Table 1). Increasing the concentration of exposure led to increased levels of penetration into PMNs, but the increase was not proportional with the concentration. The ratios of sanfetrinem were similar to those observed for both pefloxacin (10) and roxithromycin (8); in contrast, cefazolin, like other β-lactam antibiotics that we have studied (4), penetrated phagocytic cells poorly: the intracellular concentration of this drug was much lower than the extracellular level (C/E ratio, ≤0.3). The uptake of sanfetrinem proceeded rapidly, was essentially complete within 5 min, and was not energy dependent since it was not influenced by cell viability, physiologic environmental temperature, or the addition of a metabolic inhibitor (Table 2). Sanfetrinem appeared to have a great capacity to bind to intracellular proteins. In fact, an analysis of the free drug/bound drug ratios indicates that many binding sites were available for sanfetrinem, suggesting that PMNs charged with the antibiotic can carry it and release it at the site of infection, thus representing a useful biological delivery system (2). However, a documented intracellular penetration of antibiotic does not prove that it will be effective; to be clinically useful, it must retain its antimicrobial activity and not have negative effects on subsequent phagocyte functions, mainly phagocytosis or killing. In fact, some antibiotics, despite their known uptake by phagocytes, are found not to be active intracellularly (19, 36). Hence, in the next series of experiments, we examined the consequences of penetration of sanfetrinem into human PMNs on both the consequent phagocytosis and the intracellular killing of a penicillin-resistant strain of S. pneumoniae, a human pathogen whose incidence is increasing at an alarming rate worldwide, limiting the number of potentially adequate treatment regimens (18, 27, 28, 31).

When bacteria, phagocytes, and antibiotics are coincubated, a general absence of phagocyte function, impairment, and synergistic activity with phagocytes has been shown for most β-lactam antibiotics (4, 7, 9, 11, 17, 33). In contrast, sanfetrinem at a concentration of one-half the MIC was able to enhance either the bacterial uptake or the intracellular bactericidal activity of phagocytosed penicillin-resistant S. pneumoniae by human PMNs over all three time points tested compared with the results for the controls (Table 3). The mechanism of such enhancement is still unknown, although direct damage to the bacterium by the antibiotic may, at least in part, be responsible. Sanfetrinem acts by binding primarily to PBP 2 and PBP 4 of gram-positive bacteria (16) and hence exerts a profound effect on the ultrastructural morphology of S. pneumoniae, resulting in changes in the cell surface that may alter bacterial susceptibilities to PMNs functions.

The bactericidal activity of serum and the phagocytic capacities of PMNs constitute important host defense mechanisms against invading bacteria; however, our results indicated that preopsonization of S. pneumoniae resulted in no significant differences in PMN uptake and bacterial killing; furthermore, no synergistic effect between sanfetrinem and serum activity has been detected (data not shown). These findings are strongly related to the presence in the S. pneumoniae serotype 19 strain used in this study of a capsule which renders the bacterium highly resistant to the opsonization.

In order to differentiate between the effects of sanfetrinem on penicillin-resistant S. pneumoniae from those on the granulocytes, the phagocytic and bactericidal activities of PMNs against streptococci were assessed following preexposure of PMNs and bacteria individually to one-half the MIC of sanfetrinem. Once sanfetrinem had been accumulated intracellularly, it could effectively exert its bactericidal action toward the ingested bacteria. In fact, following preexposure of human PMNs to one-half the MIC of the drug, there was a significant enhancement of intracellular bactericidal activity of phagocytosed streptococci compared with the activities of the controls (Table 4). It has been reported that preexposure of phagocytes to antibiotics impairs phagocytosis (26). In contrast, our results show that preincubation of PMNs with sanfetrinem caused the streptococci to be more efficiently phagocytosed in comparison with the efficiency of phagocytosis in antibiotic-free systems. These results may suggest that sanfetrinem might act directly on the phagocyte, possibly by interfering with cellular membrane functions and hence enhancing the engulfment of the bacteria (Table 4).

Preexposure of S. pneumoniae to one-half the MIC of sanfetrinem modified the interaction between streptococci and granulocytes: the pretreated bacteria were ingested at the same rate as untreated streptococci but were more susceptible to the microbicidal intracellular mechanisms of human PMNs compared with the susceptibilities of bacteria which had not undergone previous exposure (Table 5). The mechanism of this sensitization is not clear, but it is probably due to some direct sanfetrinem-induced morphological alterations upon S. pneumoniae itself; these changes may not facilitate bacterial uptake by PMNs but, alternatively, could make the killing of penicillin-resistant S. pneumoniae internalized by PMNs easier.

In conclusion, the results of the current study indicate that sanfetrinem is able both to rapidly and efficiently penetrate human PMNs in its microbiologically active form by a passive process and to enhance directly the PMN functions, modifying the interaction between penicillin-resistant S. pneumoniae and phagocytes. Moreover, once it had been accumulated, sanfetrinem is capable of acting effectively on the replicating intraphagocytosed streptococci. Therefore, the penetration of sanfetrinem into phagocytic cells in collaboration with cellular bactericidal mechanisms, combined with its particularly broad spectrum of activity, its high potency, its resistance to β-lactamases, and its complete stability to dehydropeptidases, might have relevance for the use of this antibiotic for the treatment of a wide range of infections, including those caused by intraphagocytic pathogens, particularly in immune system-compromised or severely debilitated patients.