CD52-Negative NK Cells Are Abundant in the Liver and Less Susceptible to Alemtuzumab Treatment

Abstract

Background: T-cell depleting strategies have become an integral part of immunosuppressive regimens in organ transplantation. Alemtuzumab is a humanized monoclonal antibody against CD52, a cell-surface antigen on several immune cells. It has been suggested that lymphocyte depletion increases the risk of serious infections. However, this has not been observed with short-term alemtuzumab treatment in an organ transplant setting. For induction therapy using alemtuzumab following liver transplantation, we found that T- and B-cell numbers declined rapidly after alemtuzumab therapy; however, the natural killer (NK) cell number was sustained. NK cells are important effectors of innate immunity. Since the effects of alemtuzumab on NK cell functions, especially those of liver NK cells, are unknown, this study aimed to investigate this in detail.

Methods: To assess the effect of alemtuzumab on NK cells, samples were obtained from 7 organ donors and examined by flow cytometry using Annexin V and propidium iodide. Phenotypical and functional differences within subsets of NK cells with different levels of CD52 expression were determined by flow cytometry and in vitro cytotoxicity assays.

Results: CD52 expression on NK cells was lower than that on other lymphocyte subsets. The liver contained a large number of CD52- NK cells compared with the peripheral blood. In vitro treatment of liver-derived NK cells with alemtuzumab did not result in cell death. In contrast, co-incubation with alemtuzumab induced cell death in peripheral blood mononuclear cells and non-NK cells in the liver. Furthermore, CD52- liver NK cells were more cytotoxic and produced more IFN-γ than CD52+ NK cells after cytokine activation.

Conclusion: The liver contains a large number of CD52- NK cells. These cells are refractory to alemtuzumab and have robust activity. These findings indicate that CD52- NK cells persist and could protect against infection after alemtuzumab-based lymphocyte depletion.

Fig 1. The liver contained a high percentage of CD52⁻ NK cells.

CD52⁻ and CD52⁺ NK cells were defined as follows: Cells were gated on the CD3⁻CD56⁺ NK cell, CD3⁻CD56^bright, or CD3⁻CD56^dim populations within singlet and lymphocyte gates. A CD52⁻ gate was set using the isotype control. Data are representative of 7 separate experiments. (A) CD52 expression on LMNCs and PBMCs. CD52 expression levels on LMNCs and PBMCs were evaluated by FCM. LMNCs contain a significantly larger proportion of CD52⁻ cells when compared with PBMCs. The bar graph shows the mean ± SD of CD52⁻ lymphocytes in the liver and peripheral blood (n = 7, *p < 0.05 by Student’s t-test) (B) CD52 expression on NK cells in the liver and peripheral blood. CD52 expression on NK cells from the liver was significantly lower than that on NK cells from the peripheral blood. The bar graph shows the mean ± SD of CD52⁻ NK cells derived from LMNCs and PBMCs (n = 7, *p < 0.05 by Student’s t-test) (C) CD52 expression on liver CD56^bright and liver CD56^dim NK cells. CD52 expression levels of each population were calculated by FCM. About 90% of CD56^bright liver NK cells (10.2% ± 5.7) did not express CD52. CD52 expression on CD56^dim NK cells was significantly higher when compared with that on CD56^bright liver NK cells. The bar graph shows the mean ± SD of CD52⁻ cells in each NK cell population (n = 7, *p < 0.05 by Student’s t-test).

Fig 2. Effect of alemtuzumab concentration on survival of NK cells from the liver and peripheral blood.

Cell survival was evaluated using Annexin V and propidium iodide staining. Mononuclear cells were incubated alone or in the presence of alemtuzumab (100, 10, 1, or 0.1 μg/mL) for different time periods (1 and 4 h). The surviving cells were negative for both Annexin V and propidium iodide. (A) The survival rate of LMNCs after co-incubation with alemtuzumab was significantly higher than that of PBMCs for each concentration. The numbers present the proportion of each subset. The dot plots are representative of 7 independent experiments (alemtuzumab; 100 μg/mL, 4-h culture). The bar graphs show the mean survival rate ± SD of LMNCs and PBMCs after 4-h treatment at each concentration (n = 7, *p < 0.05 by Student’s t-test). (B) The survival rate of liver NK cells was significantly higher than that of liver-derived non-NK cells (n = 4, *p < 0.05 by Student’s t-test). (C) CD56^bright NK cells survived after co-incubation with alemtuzumab, compared to CD56^dim NK cells (n = 4, *p < 0.05 by Student’s t-test).

Fig 3. CD52⁺ and CD52⁻ NK cells from the liver had different FACS profiles.

(A) Comparison of CD52⁺ NK cells and CD52⁻ NK cells in the liver. The representative histograms of 7 independent experiments are shown for CD52⁺ NK cells (solid line) and CD52⁻ NK cells (dotted line). The gray solid line shows the isotype control. (B) Dot shows the percentage of each surface marker on CD52^- and CD52⁺ cells. The solid line indicates mean value in each population and two points connected by dotted line indicate these cells are from same donor. (*p < 0.05 by Student’s paired t-test). (n = 4 or 7, *p < 0.05).

Fig 4. Strong cytotoxicity and IFN-γ production of CD52⁻ NK cells.

(A) The CD52 sorting strategy is shown. NK cells were pre-sorted from LMNCs using an MACS NK cell isolation kit and separated into CD52 positive and negative fractions using an FACS-Aria instrument. (B) Cytotoxicity toward K562 target cells was analyzed using a flow cytometry-based cytotoxic assay. The CD52 positive and negative NK cell fractions were stimulated with IL-2 (1000 U/mL) for 48 h. The effector to target ratio was 10:1. Data are presented as the mean ± SD (n = 4, *p < 0.05 by Student’s t-test). (C) The IFN-γ level of the culture supernatant in Fig 3B was measured by ELISA. Data are presented as the mean ± SD (n = 4, *p < 0.05 by Student’s t-test). (D) The IFN-γ production of both the CD52-positive and -negative NK cell fractions was analyzed by intracellular flow cytometry staining. Freshly isolated LMNCs were treated with leukocyte activation cocktail for 4 h. The numbers indicate the IFN-γ-positive percentage of each population. Data are presented as the mean ± SD (n = 4, *p < 0.05 by Student’s t-test).