Application of atomic force microscopy (AFM) to biological objects and processes under physiological conditions has been hampered so far by the deformation and destruction of the soft biological materials invoked. Here we describe a new mode of operation in which the standard V-shaped silicon nitride cantilever is oscillated under liquid and damped by the interaction between AFM tip and sample surface. Because of the viscoelastic behavior of the cellular surface, cells effectively "harden" under such a tapping motion at high frequencies and become less susceptible to deformation. Images obtained in this way primarily reveal the surface structure of the cell. It is now possible to study physiological processes, such as cell growth, with a minimal level of perturbation and high spatial resolution (approximately 20 nm).

Changes in organization of F‐actin during the cytotoxic process between NK and K562 cells have been observed and studied using confpcal laser scanning microscopy and quantitative fluorescence microscopy. An increase in F‐actin content and orientation of F‐actin towards the target cell have been observed in, conjugated NK cells. The increase in F‐actin content probably reflects activation of the NK cell for the killing process. An increase in F‐actin content in the conjugated K562 cell, occurring simultaneously with the appearance of filamentous actin structures that often originated/ended at the contact place with the NK cell, was also observed. These changes were delayed compared to the increase in F‐actin content in the NK cell and were accompanied by increasing cytotoxic activity. This indicates that they were results of the interaction of the K562 cell with the activated NK cell. The possible role of target cell microfilaments in the cytotoxic process is addressed. © 1994 Wiley‐Liss, Inc.

The cytotoxic interaction between cloned human Natural Killer (NK) cells and K562 target cells was studied using confocal laser scanning microscopy (CUM) and conventional fluorescence microscopy. We observed, using fixed as well as living cells, the occurrence of (pseudo)emperipolesis during the interaction. About 30% of conjugated NK cells penetrated, partly or completely, into the target cells (in-conjugation). Virtually all in-conjugated target cells exhibited polymerized actin. Killer cells of inconjugates were frequently seen approaching the target cell nucleus or aligning along it. If the cytotoxic process was inhibited by the absence of calcium neither actin polymerization nor in-conjugation were observed. A kinetic study showed that in-conjugation starts somewhat later than actin polymerization but still within a few minutes after addition of calcium to conjugates previously formed Elucidation of the killing mechanism(s) involved in cell mediated cytotoxic processes has been the subject of numerous studies (6,9,16,23,32). In spite of great efforts of different research groups, the exact cause of target cell death has not been resolved yet. In general, two hypotheses dominate the cytotoxicity literature. According to the "colloid osmosis" mechanism, target cells die due to an osmotic imbalancehhock caused by incorporation of killer cell granule material into the target cell membrane (9,3032). The "internal disintegration" model, on the other hand, implies that interaction with a killer cell activates the self-destroying pathway in the target cell which results in DNA fragmentation and cell death (6,16,23). These two mechanisms, however, need not to be mutually exclusive (33). It is also not clear how much each of them participates in the actual killing processand/or what is the crucial change that ultimately leads to target cell death. It remains to be elucidated whether the changes observed in the target cell during the cytotoxic process, such as pore formation and DNA fragmentation, represent the ultimate cause or simply the consequence of approaching cell death.in the absence of calcium. The presence of cytochalasin D (an inhibitor of actin polymerization) completely inhibited in-conjugation and partly reduced the cytotoxic activity. Zinc ions (endonuclease inhibition) inhibited in-conjugation and decreased the total number of target cells with polymerized actin in a concentration dependent manner. Cytotoxic activity was also reduced but not as efficiently as in-conjugation.Our study demonstrates that in-conjugation represents a significant fraction of the cytotoxic interaction. The results indicate that it may be a consequence of an actin polymerization and endonuclease activity dependent part of a cytotoxic mechanism. o 1995 wi~ey-~iss, ~n c .Key terms: Natural Killer cell, emperipolesis, actin polymerization, endonuclease activity One of the crucial steps during cell mediated cytotoxic interaction is formation of a close contact (conjugate formation) between killer and target cells (1,9,28). The att...

Summary. The occurrence of cytotoxic lymphocyte subpopulations (i.e., CD 16 ÷, CD57 ÷ and cytotoxic CD 8 ÷) was studied in the peripheral blood of 18 B-cell chronic lymphocytic leukemia (B-CLL) patients. The absolute numbers of CD 57 ÷, CD 16 ÷ and cytotoxic CD 8 ÷ lymphocytes were increased in the peripheral blood of untreated patients as compared with healthy donors, suggesting a causal relation with the accumulation of malignant B-cells. For 5 B-CLL patients and 5 hematological normal donors, the lymphocyte subpopulations in peripheral blood, lymph nodes and bone marrow were determined. A significant immune response was observed in the lymph nodes of the patients, as reflected by the CD 3 ÷ lymphocytes, which were 1.7 -2 7 times larger in the patients lymph nodes than in their peripheral blood and bone marrow. In contrast, with peripheral blood this was mainly caused by an increase in CD 4 + lymphocytes. The CD 57 lymphocytes in the lymph nodes of the patients had abnormal orthogonal light-scattering signals and an abnormal density of CD 57 ÷ receptors in comparison with their peripheral blood CD 57 ÷ lymphocytes or the CD57 ÷ lymphocytes in the peripheral blood, bone marrow and tonsils of the hematological normal donors. This study shows that although a significant increase of cytotoxic lymphocytes in the peripheral blood of B-CLL patients is observed, the actual distributions of the non-malignant lymphocytes can be quite different at the actual t u m o r sites, i.e., bone marrow and lymph nodes.

A flow cytometer using a solid state light source and detector was designed and built. For illumination of the sample stream two types of diode lasers (670 nm and 780 nm) were tested in a set-up designed to differentiate human leukocytes by means of light scattering. The detector is an avalanche photodiode, which was used to detect the weak scattered light in the orthogonal direction. The new flow cytometer set-up is very small, relatively cheap and yields similar results as a standard flow cytometer set-up using a helium-neon laser and photomultipliers.0 1993 Wiley-Liss, Inc.