Impaired Autophagy in Adult Bone Marrow CD34+ Cells of Patients with Aplastic Anemia: Possible Pathogenic Significance

Abstract

Aplastic anemia (AA) is a bone marrow failure syndrome that is caused largely by profound quantitative and qualitative defects of hematopoietic stem and progenitor cells. However, the mechanisms underlying these defects remain unclear. Under conditions of stress, autophagy acts as a protective mechanism for cells. We therefore postulated that autophagy in CD34+ hematopoietic progenitor cells (HPCs) from AA patients might be impaired and play a role in the pathogenesis of AA. To test this hypothesis, we tested autophagy in CD34+ cells from AA samples and healthy controls and investigated the effect of autophagy on the survival of adult human bone marrow CD34+ cells. We found that the level of autophagy in CD34+ cells from AA patients was significantly lower than in age/sex-matched healthy controls, and lower in cases of severe AA than in those with non-severe AA. Autophagy in CD34+ cells improved upon amelioration of AA but, compared to healthy controls, was still significantly reduced even in AA patients who had achieved a complete, long-term response. We also showed that although the basal autophagy in CD34+ cells was low, the autophagic response of CD34+ cells to "adversity" was rapid. Finally, impaired autophagy resulted in reduced differentiation and proliferation of CD34+ cells and sensitized them to death and apoptosis. Thus, our results confirm that autophagy in CD34+ cells from AA patients is impaired, that autophagy is required for the survival of CD34+ cells, and that impaired autophagy in CD34+ HPCs may play an important role in the pathogenesis of AA.

Fig 1. Impaired autophagy in CD34⁺ cells from AA patients.

(A) and (E): NanoPro1000^TM system analysis. Samples were analyzed immediately after isolation without being first cultured. The peaks at pI 5.70, pI 8.36 and pI 9.95 represent LC3-I, LC3-II and GAPDH, respectively. The ratio of the peak area of LC3-II normalized to GAPDH from the same sample is presented, demonstrating that healthy controls had higher LC3-I/LC3-II than NSAA/SAA (SAA group, n = 7; NSAA group, n = 9; control group, n = 4). (B) and (F): NanoPro1000^TM system analysis. CD34⁺ cells were cultured in the presence of 20 μM CQ for 24 h prior to analysis. The addition of CQ accumulated LC3-II expression in all groups, though healthy controls had higher LC3-I/LC3-II than did AA (AA group, n = 7; control group, n = 3). (C) and (G): Confocal microscopy (×1,000). Samples were analyzed immediately after isolation without being first cultured. The LC3B fluorescence density of each sample is shown. Healthy controls exhibited punctate LC3B (red), NSAA/SAA showed a weak cytosolic LC3B distribution, and the LC3B fluorescence density in the NSAA/SAA group was significantly lower than in the control group (SAA group, n = 5; NSAA group, n = 5; control group, n = 3). (D) and (H): Confocal microscopy (×1,000). Samples were analyzed immediately after isolation without being first cultured. The MDC fluorescence density of each sample is shown. Healthy controls exhibited more MDC-labeled particles (blue) and MDC fluorescence density than did the NSAA/SAA (SAA group, n = 5; NSAA group, n = 5; control group, n = 3). * P<0.05, ** P<0.01, *** P<0.001.

Fig 2. Improved autophagy in CD34⁺ cells from AA patients with the amelioration of disease.

All freshly sorted CD34⁺ cells were analyzed immediately after isolation without culture. (A) and (D): NanoPro1000^TM system. The LC3-I and LC3-II peak and the ratio of the LC3-II peak area normalized to GAPDH of each sample is presented (NR AA group, n = 9; PR AA group, n = 7; CR AA group, n = 7; control group, n = 4). With the amelioration of AA, the level of LC3-II increased but was still lower than in healthy controls. (B) and (E): Confocal microscopy (×1,000). The LC3B fluorescence and mean LC3B fluorescence density of CD34⁺ cells is shown (NR AA group, n = 5; PR AA group, n = 7; CR AA group, n = 5; control group, n = 3). With the amelioration of AA, the mean LC3B fluorescence density increased, but was still lower than in healthy controls. (C) and (F): Confocal microscopy (×1,000). The MDC fluorescence and mean MDC fluorescence density of CD34⁺ cells is shown (NR AA group, n = 5; PR AA group, n = 6; CR AA group, n = 7; control group, n = 3). With the amelioration of AA, the mean MDC fluorescence density increased, but was still lower than in healthy controls. * P<0.05, ** P<0.01, *** P<0.001.

Fig 3. Autophagic heterogeneity in CD34⁺ cells from AA patients.

(A): LC3B staining imaged with confocal microscopy (×200). A few residual CD34⁺ cells from AA patients showed a punctuate immunostaining pattern, just like CD34⁺ cells from healthy controls. (B): MDC staining by confocal microscopy (×200). A few residual CD34⁺ cells from AA patients contained MDC-labeled particles, just like CD34⁺ cells from healthy controls.

Fig 4. Freshly sorted normal adult human bone marrow CD34⁺ cells had higher autophagic activity than lymphocytes and neutrophils.

(A) and (D): TEM (×10,000). Autophagosomes and the mean number of vacuoles (autophagosomes) in CD34⁺ cells, lymphocytes and neutrophils. CD34⁺ cells contain more vacuoles (autophagosomes) than lymphocytes and neutrophils. In each cell population, the mean number of vacuoles per cell (mean±SEM) is presented. (B) and (E): Confocal microscopy (×1,000). LC3B fluorescence and mean LC3B fluorescence density is shown. CD34⁺ cells exhibit higher mean fluorescence density than lymphocytes and neutrophils. (C) and (F): NanoPro1000^TM system. The LC3-I and LC3-II peaks and the ratio of the LC3-II peak area normalized to GAPDH is shown. CD34⁺ cells have higher levels of LC3-II than lymphocytes and neutrophils.

Fig 5. Levels of autophagy in healthy adult bone marrow CD34⁺ cells at different time-points of nutrient deprivation.

Freshly sorted CD34⁺ cells were cultured for 12 h and placed in PBS for 0 min, 30 min, 60 min and 120 min after harvest. (A) and (C): Analysis of LC3 protein using the NanoPro1000^TM system. The LC3-I and LC3-II peaks and the ratio of LC3-II peak area normalized to GAPDH at different time-points is shown. The LC3-II peak and peak area rapidly increased with the duration of nutrient deprivation. (B) and (D): Analysis of LC3B using confocal microscopy. LC3B fluorescence and mean LC3B fluorescence density is presented. The LC3B fluorescence density in CD34⁺ cells was weak at 0 min and rapidly increased with the duration of nutrient deprivation.

Fig 6. Autophagy effects on the survival of CD34⁺ cells.

(A) and (B): CD34⁺ cells were treated with 3-MA or CQ, and the number of cells was counted at days 4, 7 and 10. 3-MA and CQ were able to inhibit proliferation of CD34⁺ cells in a concentration-dependent manner. The results shown are the mean±SEM of three independent experiments. (C) and (D): CFC assay. CD34⁺ cells were treated with 3-MA or CQ, and colonies were counted at day 10 and normalized to control conditions. 3-MA and CQ inhibited differentiation of CD34⁺ cells in a concentration-dependent manner. The results shown are the mean±SEM of three independent experiments. (E) and (F): Trypan blue assays revealing the percentage of dead cells. CD34⁺ cells were treated with 3-MA or CQ for 48 h prior to application of trypan blue. 3-MA and CQ induced the death of CD34⁺ cells in a concentration-dependent manner. The data represent the mean±SEM of three independent experiments. (G-I): Apoptosis detection by AnnexinV-FITC/PI assays. CD34⁺ cells were treated with 3-MA or CQ for 24 h, followed by staining with Annexin-V/PI. 3-MA and CQ induced the apoptosis of CD34⁺ cells in a concentration-dependent manner. G and H represent the mean±SEM of three independent experiments, with the Q2 quadrant (Annexin V+/PI+) and Q3 quadrant (Annexin V+/PI-) in I representing the percentages of early and late apoptotic cells, respectively. * P<0.05, ** P<0.01, *** P<0.001.