Main

Leukocytes must deform markedly to pass through narrow capillaries with diameters of about 5 µm (1). Both neonatal PMN and erythrocytes have resting diameters of approximately 9 µm (2,3), but mature PMN have a 4 times larger volume (2,3) and 500 times higher flow resistance through 5-µm filter pores (4) than do erythrocytes. Fortunately, leukocytes usually travel through relatively wide capillaries with high flow, thereby requiring little or no deformation (5). At low perfusion pressure, however, leukocytes may enter into narrow capillaries with low flow and obstruct these vessels, thereby further deteriorating flow through ischemic tissues. A major role of PMN accumulation in the development of severe tissue injury has been shown in adult animal models with cardiac (6), cerebral (7), and gastrointestinal (8) ischemia. Capillary leukostasis may also contribute to the "no-reflow" phenomenon after local or general ischemia (9). Recently, PMN accumulation has been shown in brain vessels of the neonatal rat after ligation of the right common carotid artery and systemic hypoxia (10).

Accumulation of leukocytes in low flow tissues is intensified by increased cell rigidity and large cell volume (5,11). Immature granulocytes in neonates and adults are both less deformable and larger than are mature PMN (2,12–14). In neonatal septicemia, up to 100% of the circulating neutrophils may be immature (15). In addition to their impaired function (16), the rigid and large immature neutrophils may more than mature PMN tend to obstruct narrow vessels.

In neonates with septicemia (17) and severe surfactant deficiency (18), a large percentage of neutrophils may become activated. Active PMN are very rigid both in vitro (19) and in vivo (1,20) and may therefore rapidly obstruct narrow vessels (1,6,20).

Previous studies have shown that erythrocyte deformability may markedly decrease in Gram-negative (21) and Gram-positive septicemia (22). Moreover, in adults with septicemia the passage time of isolated PMN may be prolonged by 100-200% (23). Because individual PMN have not been studied in septicemia, it is unclear whether impaired deformability of PMN is a result of increased numbers of rigid immature or activated neutrophils or of increased overall rigidity of the circulating neutrophils. Moreover, no study on the deformability of neutrophils in neonates with septicemia appears to exist in the literature. The present investigation was designed to compare the volume and deformability of passive and active mature PMN and of immature neutrophils in healthy and septicemic neonates.

METHODS

Patients. Six groups (see Table 1) were studied with approval of the ethics committee of the University of Heidelberg Medical Center: 1) 20 healthy nonsmoking adults; 2) 20 healthy full-term infants; 3) 20 healthy preterm infants without signs of infection, respiratory distress, or other disorders and without need of assisted ventilation (28-32 wk of gestation); 4) 9 neonates with GBS septicemia (33-40 wk of gestation); 5) 10 neonates with Gram-negative septicemia (6 Escherichia coli, 2 Enterobacter cloacae; 2 Klebsiella oxytoca; 28-37 wk of gestation); and 6) 7 neonates with Staphylococcus epidermidis septicemia (28-32 wk of gestation). All septic babies showed clinical signs of septicemia, CRP > 20 mg/L (nephelometry), and septicemia was proven by a positive blood culture. The blood samples for the micropipette studies (0.5 mL) were anticoagulated with EDTA (1 mg/mL) (18). All studies on one blood sample were finished within 3 h. The blood samples of the neonates were primarily taken for routine laboratory tests (blood cultures, CRP, leukocyte count, differential blood picture, screening tests, and so forth). None of the infants and adults received antibiotics at the time of blood sampling. Informed consent of the parents was obtained before blood sampling.

Table 1 Leukocyte indices and CRP in healthy and septicemic humans
Full size table

Micropipette system. Pipettes with internal diameters of 2.5 µm were used to aspirate membrane-cytoplasm tongues of granulocytes (24–26). Details of the leukocyte preparation, the micropipette system, and the procedure of leukocyte deformation studies have been described elsewhere (2,14). The blood was centrifuged at 350 × g and 4°C for 25 min. Leukocytes were taken directly from the buffy coat. A separation medium was not used. The cells were observed using a light microscope with an optical magnification of X788. The additional electronic magnification of the video system resulted in a final magnification of X4000. The studies were carried out at 22°C. An aspiration pressure of 2 cm H2O was applied. The length of the aspirated tongue was measured from video recording after 5, 10, 20, 30, 40, 50, and 60 s. The diameter (D) of each cell was measured before aspiration. Because passive neutrophils had a spherical shape (Fig. 1), the volume (V) of each individual cell was calculated as V = ([pi]a/6)(D3). Volume of active PMN was calculated only if the cell shape was spherical.

Figure 1
figure 1

Passive (left panel) and active (right panel) PMN. The active PMN shows several pseudopods.

Full size image

Leukocyte subpopulations. The leukocyte subpopulations were identified during the microscopic deformation studies. Several previous studies have shown that different types of leukocytes, of neutrophils and their precursors as well as activated neutrophils can be distinguished by light microscopy (2,12,14,24–27). Sometimes the granules interfere with the view of the nucleus. But the nucleus always becomes clearly visible upon tongue aspiration (25). Mature neutrophils were distinguished from eosinophils and basophils by their smaller and lighter granules. Mature and immature neutrophils were distinguished by the shape of their nucleus. Myeloblasts and promyelocytes contain a round or oval nucleus. Myelocytes contain an oval or indented nucleus. Because of their small number, we combined the results of myeloblasts, promyelocytes, and myelocytes. Metamyelocytes were defined as cells with a nucleus shaped like a kidney or a rod. In band cells, the nucleus is of uniform thickness and curved throughout. In mature PMN, the nucleus consists of two to five or more distinct lobes.

Activation of neutrophils was assumed if pseudopod projections greater than 0.5 µm were visible (Fig. 1), tongue growth in the pipette abruptly slowed or stopped, or tongues partially retracted from the pipette during aspiration (25–27).

In each blood sample, the distribution of passive and activated mature PMN and of immature neutrophils were initially counted among 100 PMN. Then cells were purposely selected for the deformation studies to yield sufficient numbers for each cell subpopulation. In each sample, 20 passive mature PMN were studied. PMN activated upon contact with the glass surface of the micropipette were added to the number of activated cells showing pseudopods before aspiration into the pipette.

Statistical analysis. Means and SD were calculated for leukocyte counts, percentages of neutrophil subpopulations and CRP (Table 1), and for the neutrophil volumes and tongue lengths (Table 2). Table 2 presents the values for the entire cell populations of each cell type for each of the six groups. Statistical analyses were performed to test for intergroup differences between the combined values of the healthy full-term and preterm infants and each of the three groups of septicemic infants. Moreover, differences of volume and deformation behavior among various neutrophil subpopulations were analyzed separately for each of the six groups (Table 2). Kruskal-Wallis nonparametric analysis of variance was used to calculate significances. Because of the large number of comparisons, significant differences were defined as p < 0.005.

Table 2 Volume and tongue growth of mature and immature neutrophils in healthy and septic humans*
Full size table

RESULTS

The total leukocyte and neutrophil counts, the percentages of the neutrophil subpopulations, and the CRP values are summarized in Table 1. Neutropenia (neutrophil count <2 × 109/L) was observed in 3 of 9 infants with GBS, 4 of 10 with Gram-negative, and none with S. epidermidis septicemia. One infant with E. coli and two with S. epidermidis septicemia had neutrophilia (neutrophil count >25 × 109/L).

Cellular volume increased progressively with decreasing leukocyte maturity. The volume of each cell type was similar in the six groups and not affected by septicemia (Table 2). Compared with mature PMN, the volumes of immature neutrophils were increased by 14-22% (band cells), 28-36% (metamyelocytes), and 36-42% (myelocytes, promyelocytes, and myeloblasts).

All passive cells showed steady tongue formation in 2.5-µm micropipettes (Table 2; Fig. 2). Immature neutrophils showed diminished tongue growth. Compared with mature passive PMN, final tongue lengths of immature neutrophils were decreased by about 50% (band cells), 60% (metamyelocytes), and 70% (myelocytes, promyelocytes, myeloblasts). Tongue growth of the same cell type was similar in neonates and adults and was not significantly affected by septicemia (Table 2).

Figure 2
figure 2

Time-dependent membrane cytoplasm tongue formation of various neutrophil types in 2.5-µm micropipettes. Shown are the mean values obtained in nine neonates with GBS septicemia.

Full size image

In infants with septicemia, a substantial number of the mature neutrophils were active or became activated upon aspiration into glass pipettes. Neonates with GBS (14%) and Gram-negative septicemia (12%) showed more active PMN than infants with S. epidermidis septicemia (8%). Immature neutrophils showed no activation.

We observed three types of active neutrophils among the 88 active PMN studied in septicemic infants and the 56 active PMN in healthy blood donors: 1) cells with visible pseudopods before aspiration of a membrane-cytoplasm tongue into the pipette (32 and 11%, respectively); 2) cells with abrupt stoppage or slowing of tongue growth (42 and 64%), and 3) retraction from the pipette in spite of continuous aspiration (26 and 25%). Most active PMN showed relatively fast tongue growth during the first 5 s, but little or no further tongue growth during the following 55 s (Fig. 3).

Figure 3
figure 3

Time-dependent membrane cytoplasm tongue formation of active PMN. Shown are the mean values for the three types of active PMN studied in the 26 septic neonates. Pseudopods were studied in 20 PMN.

Full size image

The main cell body of PMN with visible activation before aspiration showed similar tongue formation as passive cells during the first 5 s of aspiration. Thereafter, the tongue growth was markedly reduced. Cells activated during aspiration demonstrated a sudden cessation of tongue growth at the time of activation. Passive deformation of pseudopods was studied in 20 PMN from septic neonates. Pseudopods formed small tongues only (Fig. 3), suggesting marked rigidity.

DISCUSSION

During bacterial infection, large numbers of neutrophils are released from the bone marrow into the circulating blood (15). Many of these cells are rapidly sequestered in the vascular bed of infected tissues. If the supply is less than the sequestration, blood neutropenia results. The increase in neutrophil release from the bone marrow is usually associated with the release of immature neutrophils. In our study, one third of the circulating PMN were immature in neonates with GBS and Gram-negative septicemia, whereas in neonates with S. epidermidis septicemia only 17% of the PMN were immature (Table 1).

During septicemia, several inflammatory mediators are activated or released from the infected tissues and cause activation of both blood vessel endothelium and leukocytes (28). Neutrophils are recruited for phagocytosis and microbial killing at the site of infection or injury. The expression of adhesion molecules and their ligands increases, thereby causing local accumulation and adhesion of leukocytes. Due to their elevated adhesion energy, active neutrophils are usually not found in circulating blood (27). However, we observed that in vitro manipulation may cause visible activation of a substantial number of neutrophils in septicemia (Table 1). This suggests that in septicemia circulating neutrophils are more susceptible to activation.

Neutrophils activated in vitro or in vivo with chemoattractants are much stiffer than passive neutrophils (1,19–20). In particular, pseudopods formed during activation are extremely rigid (29). We have shown similar behavior of active PMN and their pseudopods in septic neonates. Pseudopods are formed as a result of polymerization and alignment of actin molecules near the PMN membrane surface. The rigidification of pseudopods may be explained by the change of the physical properties of the membrane cytoplasm from a viscoelastic semifluid to an elastic solid and the loss of excess membrane surface area (29). Because the activated peripheral neutrophils behaved similarly as neutrophils with chemotactic activation (11,20), we assume that the activated peripheral neutrophils are representative of neutrophils that are sequestrated in vivo.

Immature neutrophils are larger and less deformable than mature neutrophils. In spite of their poor deformability, even high counts of immature leukocytes (e.g. in leukemia) rarely cause small vessel obstruction (leukostasis) and organ ischemia, as long as the perfusion pressure is normal (30). In vessels with diameters of less than 200 µm, the large leukocytes travel in the rapid center stream and preferentially flow into relatively wide branches with high flow. Upon reduction of flow, leukocytes are displaced from the center stream to the slowly flowing vessel margin and travel into relatively narrow branches with slow flow, thereby further deteriorating blood flow in these vessels (5). The extent of capillary leukostasis at low perfusion pressure depends on the deformability of the leukocytes (11). Thus, although immature neutrophils show little or no motility and adhesion (16), they may enhance leukostasis and capillary obstruction at low perfusion pressure.

We conclude that increased stiffness of immature and activated leukocytes may play an important role in the development of vessel obstruction in septicemia; immature neutrophils are larger and more rigid than mature PMN and may, therefore, intensify leukocyte sequestration in septicemia. Activated PMN are extremely rigid and readily adhere to endothelium, thereby obstructing small vessels. Sequestration of PMN may contribute to pulmonary hypertension (31), ischemic brain damage (32), and necrotizing enterocolitis (8) in septicemia. We speculate that inhibition of PMN activation by drugs (e.g. pentoxifylline) (33) may ameliorate the circulatory complications of septicemia.