Learning Objectives

To obtain between 50,000–100,000 cells from FAC sorting, between 160 and 200 midguts need to be dissected. Obviously this is the most time consuming step of the whole procedure. The cartoon shown in Figure 1A illustrates the major steps of gut dissection, which are described in detail in the protocol provided here. This procedure can be individually modified. A dissected, whole gastrointestinal tract is shown in Figure 1B.

When all of the midgut tissue has been digested, immediately continue with FAC sorting. Following the gating strategy described in the protocol allows for the successful identification and sorting of a cell population enriched for ISCs and for EBs from young (Figure 2) and from old midguts (Figure 3). The P1 gate is set to include only healthy living cells and to gate out dead cells. Cells that plot above 10³ on the logarithmic y-axis when using the Pacific Blue channel are defined as dead cells. The dots that plot below 40 on the x-axis (FSC-A) are mostly debris and therefore are also excluded (Figure 2A). In the scatter plot of SSC-A vs. FSC-A the cells are distributed based on their size and granularity. For an optimal resolution, the cells are placed in the scatter plot as shown in Figure 2B by adjusting the photomultiplier tube (PMT) voltage. In the scatter plots FSC-H vs. FSC-W and SSC-H vs. SSC-W single cells are separated from aggregates and doublets based on size (Figure 2C) and based on granularity (Figure 2D). When depicting the cells included in gate P4 in a histogram plot, the distinct separation of GFP-negative, GFP^low (gate P5) and GFP^high (gate P6) cells is visible (Figure 2E). Cells exhibiting lower GFP intensity are small and less granular and represent ISCs. Cells exhibiting higher GFP intensity are larger and more granular and represent EBs. Reverse Transcriptase (RT)-PCR for Dl on cDNA synthesized from RNA of cells of each GFP-positive population (gates P5, P6) revealed that the Dl signal is indeed stronger in the smaller, GFP^low cells than in the larger, GFP^high cells (Figure 2H). The cells were then backgated and depicted in contour plots to confirm that the GFP^low (gate P5) and GFP^high (gate P6) cells differ in size and granularity (Figure 2F) and are still alive (Figure 2G). Of note, the two peaks of GFP-positive cells (GFP^low and GFP^high) can only be nicely distinguished if the FITC (GFP) channel has been calibrated properly using the controls described in the protocol.

The same gating strategy was used when sorting ISCs derived from old guts (Figure 3). The scatter plots (Figure 3A-D), the histogram (Figure 3E) and the contour plots (Figure 3F, G) are analogous to the ones shown in Figure 2. As mentioned above, there is no bona fide marker to identify ISCs in old midguts and therefore no RT-PCR data is presented here. However, the intriguing observation here is that the two peaks containing GFP^low (gate P5) or GFP^high (gate P6) cells remain distinctly separated during aging. Furthermore, the number of cells in the GFP^high (gate P6) peak significantly increases during aging, which clearly emulates the known accumulation of misdifferentiated EBs with age. The change in the ratio of the two cell populations can also be seen in the scatter plots (compare Figure 3A-D with Figure 2A-D) and in the contour plots (compare Figure 3F, G with Figure 2F, G). From these findings we conclude that by sorting for GFP-positive cells using FACS we can enrich for ISCs and EBs in young and old midguts.

To validate the purity of the isolated ISCs and EBs, a post sort analysis was performed (Figure 4). The histogram plots of the post-sorted ISCs and EBs (Figure 4B, C) depict purities between 95% and 98%.

If the hierarchy of setting the FACS parameters as described above is strictly followed, the two different cell populations (GFP^low, small and GFP^high, larger) can already be distinguished in a scatter plot GFP vs. FSC-A (Figure 5A). These two cell populations cannot be distinguished if this hierarchy is disregarded (Figure 5B-D).

Figure 1: (A) A schematic of the major steps of dissecting the fly intestine. First, the head is removed (removal of the wings is optional) followed by a separation of the thorax and abdomen to expose the gut. The gut is subsequently freed from the body cavity in the following steps. The black arrows show the progression of dissection, whereas the dark red arrows denote the pull direction. The red lines in the last image indicate the anterior and posterior boundaries of the midgut. (B) A light microscope image of the adult fly intestine. The main regions and anatomical features are labeled. Red lines mark the boundaries of the midgut.

Figure 2: FAC sorting strategy to identify and isolate ISCs and EBs from young (7 days) midguts. For better visualization, cells representing ISCs are highlighted in red and cells representing EBs are highlighted in blue in the scatter plots (A-D) and in the contour plots (F,G). (A) The P1 gate is set to exclude dead, Sytox-positive cells plotted above 10³ on the logarithmic y-axis. The threshold of the FSC-A is set at 40 to also exclude dead cells and debris. (B) Cells were gated for FSC-A and SSC-A to select cells according to size and granularity (gate P2). The histogram plot (E) was used in parallel to identify the GFP-positive cells in the FSC-A vs. SSC-A scatter plot (B). (C, D) The scatter plots FSC-H vs. FSC-W and SSC-H vs. SSC-W serve to remove aggregates and doublets and select for singlets (gate P3 in C and gate P4 in D). (E) The histogram plot shows the number of cells and their GFP intensity. Two peaks, GFP^low (gate P5) and GFP^high (gate P6) can be distinguished. (F, G) Cells were back-gated in a contour plot to verify that the two distinct GFP-positive cell populations are of the correct size (F) and contain living cells (G). (H) Reverse Transcriptase (RT)-PCR for Dl on cDNA synthesized from RNA isolated from the cells present in the first peak (gate P5) and in the second peak (gate P6), respectively. More Dl expression could be detected in cells present in gate P5 vs. gate P6. Tubulin served as a loading control. Please click here to view a larger version of this figure.

Figure 3: FAC sorting strategy to identify and isolate ISCs and EBs from old (60 days) midguts. For better visualization, cells representing ISCs are highlighted in red and cells representing EBs are highlighted in blue in the scatter plots (A-D) and in the contour plots (F,G). (A) The P1 gate is set to exclude dead, Sytox-positive cells plotted above 10³ on the logarithmic y-axis. The threshold of the FSC-A is set at 40 to also exclude dead cells and debris. Of note, the age-dependent increase in larger GFP-positive cells is already obvious in this plot. (B) Cells were gated for FSC-A and SSC-A to select cells according to size and granularity (gate P2). The histogram plot (E) was used in parallel to identify the GFP-positive cells in the FSC-A vs. SSC-A scatter plot (B). (C, D) The scatter plots FSC-H vs. FSC-W and SSC-H vs. SSC-W serve to remove aggregates and doublets and select for singlets (gate P3 in C and gate P4 in D). (E) The histogram plot shows the number of cells and their GFP intensity. Two peaks, GFP^low (gate P5) and GFP^high (gate P6) can be distinguished. Of note, the cell population GFP^high containing larger cells has increased in number during aging. (F, G) Cells were back-gated in a contour plot to verify that the two distinct GFP-positive cell populations are of the correct size (F) and contain living cells (G). Please click here to view a larger version of this figure.

Figure 4: A post sort analysis was performed to determine the purity of the isolated ISCs and EBs. (A) GFP-positive cells from young (7 days) and old (60 days) midguts were FAC sorted and the histogram plots show their distribution into two peaks based on GFP intensity. (B) The post sort analysis demonstrates the purity of the isolated ISCs from young and old midguts which is >97%. (C) The post sort analysis of sorted EBs from young and old midguts shows a purity of >95%. The slight impurities may result from fluorescence quenching or mechanically damaged GFP expressing cells caused by re-sorting the cells. Please click here to view a larger version of this figure.

Figure 5: Examples of scatter plots that show suboptimal FACS data, which is seen when the FACS parameters were not set properly. (A) Representative FACS profile of all cells when the FACS parameters were set properly based on the controls described in the protocol. The cell population containing ISCs is circled in red and the cell population containing EBs is circled in blue. The cells circled in green are dead cells that have a stronger autofluorescence compared to living cells, which are plotted below 10³ on the logarithmic y-axis. (B) The scatter plot depicts a typical result obtained when the FACS instrument was not calibrated. No GFP-positive cells are detected. (C) If the FSC is not set properly, the different GFP-positive cell populations are not detected. (D) If the FITC (GFP) channel is not adjusted correctly, the majority of GFP-positive cells to be sorted are not plotted in the scatter plot but rather along the upper edge of the plot. Also, in the plot shown here, the limit for autofluorescence was not adjusted, hence cells that plot above 10² on the logarithmic y-axis are actually autofluorescent and could be mistaken as GFP-positive cells.