Author response:
Public Reviews:
Reviewer #1 (Public review):
Summary:
This work shows that a specific adenosine deaminase protein in Dictyostelium generates the ammonia that is required for tip formation during Dictyostelium development. Cells with an insertion in the ADGF gene aggregate but do not form tips. A remarkable result, shown in several different ways, is that the ADGF mutant can be rescued by exposing the mutant to ammonia gas. The authors also describe other phenotypes of the ADGF mutant such as increased mound size, altered cAMP signalling, and abnormal cell type differentiation. It appears that the ADGF mutant has defects in the expression of a large number of genes, resulting in not only the tip defect but also the mound size, cAMP signalling, and differentiation phenotypes.
Strengths:
The data and statistics are excellent.
Weaknesses:
(1) The key weakness is understanding why the cells bother to use a diffusible gas like ammonia as a signal to form a tip and continue development.
Diffusion of a gas can affect the signalling process of the entire colony of cells and will be quicker than other signaling mechanisms. A number of findings suggest that ammonia acts as both a local and long-range regulatory signal, integrating environmental and cellular cues to coordinate multicellular development. Ammonia serves as a crucial signalling molecule, influencing both multicellular organization and differentiation in Dictyostelium (Francis, 1964; Bonner et al., 1989; Bradbury and Gross, 1989). By raising the pH of the intracellular acidic vesicles of prestalk cells (Poole and Ohkuma, 1981; Gross et al, 1983), and the cytoplasm, ammonia is known to increase the speed of chemotaxing amoebae (Siegert and Weijer 1989; Van Duijn and Inouye, 1991), triggering multicellular movement (Bonner et al., 1988, 1989) to favor tipped mound development. The slug tip is known to release ammonia while the slime sheath at the back of the slug prevents diffusion thus maintaining high ammonia levels to (Bonner et al., 1989) promote pre-spore differentiation (Newell et al., 1969). Ammonia has been found to favor slug migration rather than fruiting (Schindler and Sussman, 1977) and thus, tip-derived ammonia may stimulate synchronized development of the entire colony. The tip exerts negative chemotaxis towards ammonia, potentially directing the slugs away from each other to ensure equal spacing of fruiting bodies (Feit and Sollitto, 1987).
Ammonia released in pulses acts as a long-distance signalling molecule between colonies of yeast cells indicating depletion of nutrient resources and promoting synchronous development (Palkova et al., 1997; Palkova and Forstova, 2000). A similar mechanism may be at play to influence neighbouring Dictyostelium colonies. Furthermore, ammonia produced in millimolar concentrations (Schindler and Sussman, 1977) may also ward off predators in soil as observed in Streptomyces symbionts of leaf-cutting ants to inhibit fungal pathogens (Dhodary and Spiteller, 2021). Additionally, ammonia may be recycled into amino acids, within starving Dictyostelium cells to supporting survival and differentiation as observed in breast cancer cells (Spinelli et al., 2017). Therefore, using a diffusible gas like ammonia as a signalling molecule is likely to have bioenergetic advantages. Ammonia is a natural metabolic byproduct of amino acid catabolism and other cellular processes, making it readily available without requiring additional energy for synthesis. Instead of producing a dedicated signalling molecule, cells can exploit an existing by-product for developmental regulation.
(2) The rescue of the mutant by adding ammonia gas to the entire culture indicates that ammonia conveys no positional information within the mound.
Ammonia is known to influence rapid patterning of Dictyostelium cells confined in a restricted environment (Sawai et al., 2002). Both neutral red staining (a marker for prestalk and ALCs) (Fig. S2) and the prestalk marker ecmA/ ecmB expression (Fig. 8C) in the adgf mutants suggest that the mounds have differentiated prestalk cells but are blocked in development. The mound arrest phenotype can be reversed by exposing the adgf mutant mounds to ammonia.
Based on cell cycle phases, there exists a dichotomy of cell types, that biases cell fate to prestalk or prespore (Weeks and Weijer, 1994; Jang and Gomer, 2011). Prestalk cells are enriched in acidic vesicles, and ammonia, by raising the pH of these vesicles and the cytoplasm (Davies et al 1993; Van Duijn and Inouye 1991), plays an active role in collective cell movement (Bonner et al., 1989). Thus, ammonia reinforces or maintains the positional information by elevating cAMP levels, favouring prespore differentiation (Bradbury and Gross, 1989; Riley and Barclay, 1990; Hopper et al., 1993).
(3) By the time the cells have formed a mound, the cells have been starving for several hours, and desperately need to form a fruiting body to disperse some of themselves as spores, and thus need to form a tip no matter what.
When the adgf mutants were exposed to ammonia just after tight mound formation, tips developed within 4 h (Fig. 6). In contrast, adgf mounds not exposed to ammonia remained at the mound stage for at least 30 h. This demonstrates that starvation alone is not sufficient to drive tip development and ammonia serves as a cue that promotes the transition from mound to tipped mound formation.
Many mound arrest mutants are blocked in development and do not proceed to form fruiting bodies (Carrin et al., 1994). Furthermore, not all the mound arrest mutants tested in this study were rescued by ADA enzyme (Fig. S3 A), and they continue to stay as mounds without dispersing as spores, suggesting that mound arrest in Dictyostelium can result from multiple underlying defects, whereas ammonia is an important factor controlling transition from mound to tip formation.
(4) One can envision that the local ammonia concentration is possibly informing the mound that some minimal number of cells are present (assuming that the ammonia concentration is proportional to the number of cells), but probably even a minuscule fruiting body would be preferable to the cells compared to a mound. This latter idea could be easily explored by examining the fate of the ADGF cells in the mound - do they all form spores? Do some form spores?
Or perhaps the ADGF is secreted by only one cell type, and the resulting ammonia tells the mound that for some reason that cell type is not present in the mound, allowing some of the cells to transdifferentiate into the needed cell type. Thus, elucidating if all or some cells produce ADGF would greatly strengthen this puzzling story.
A fraction of adgf mounds form bulkier spore heads by the end of 36 h as shown in Fig. 3. This late recovery may be due to the expression of other ADA isoforms. Mixing WT and adgf mutant cell lines results in a slug with the mutants occupying the prestalk region (Fig. 9) suggesting that WT ADGF favours prespore differentiation. However, it is not clear if ADGF is secreted by a particular cell type, as adenosine can be produced by both cell types, and the activity of three other intracellular ADAs may vary between the cell types. To address whether adgf expression is cell type-specific, we will isolate prestalk and prespore cells, and thereafter examine adgf expression in each population.
ADGF activity is likely to be higher in the tip to remove excess adenosine, the tip-inhibiting molecule (Wang and Schaap, 1985). Moreover, our results show that adgf- cells with high adenosine preferentially migrate to the prestalk rather than the prespore region when mixed with WT cells. Ammonia generated from adenosine deamination could thus drive tip development and prespore differentiation.
Reviewer #2 (Public review):
Summary:
The paper describes new insights into the role of adenosine deaminase-related growth factor (ADGF), an enzyme that catalyses the breakdown of adenosine into ammonia and inosine, in tip formation during Dictyostelium development. The ADGF null mutant has a pre-tip mound arrest phenotype, which can be rescued by the external addition of ammonia. Analysis suggests that the phenotype involves changes in cAMP signalling possibly involving a histidine kinase dhkD, but details remain to be resolved.
Strengths:
The generation of an ADGF mutant showed a strong mound arrest phenotype and successful rescue by external ammonia. Characterization of significant changes in cAMP signalling components, suggesting low cAMP signalling in the mutant and identification of the histidine kinase dhkD as a possible component of the transduction pathway. Identification of a change in cell type differentiation towards prestalk fate
Weaknesses:
(1) Lack of details on the developmental time course of ADGF activity and cell type type-specific differences in ADGF expression.
ADGF expression was examined at 0, 8, 12, and 16 h (Fig. 1), and the total ADA activity was assayed at 12 and 16 h (Fig. 4). As per the reviewer’s suggestion, we have now included the 12 h data (Fig. 4A) to provide additional insights into the kinetics of ADGF activity. The adgf expression was found to be highest at 16 h and hence, the ADA assay was carried out at that time point. However, the ADA assay will not exclusively reflect ADGF activity since it reports the activity of the three other isoforms as well.
A fraction of adgf- mounds form bulkier spore heads by the end of 36 h as shown in Fig. 3. This late recovery may be due to the expression of the other ADA isoforms. Mixing WT and adgf mutant cell lines results in a slug with the mutants occupying the prestalk region (Fig. 9), suggesting that WT adgf favours prespore differentiation.
However, it’s not clear if ADGF is secreted by a particular cell type, as adenosine can be produced by both cell types, and the activity of the other three intracellular ADAs may vary between the cell types. To address whether adgf expression is cell typespecific, we will isolate prestalk and prespore cells, and thereafter examine adgf expression in each population.
ADGF activity is likely to be higher in the tip to remove excess adenosine, the tipinhibiting molecule (Wang and Schaap, 1985). Moreover, our results show that adgf- cells with high adenosine preferentially migrate to the prestalk rather than the prespore region when mixed with WT cells.
(2) The absence of measurements to show that ammonia addition to the null mutant can rescue the proposed defects in cAMP signalling.
The cAMP levels were measured at two time points 8 h and 12 h in the mutant. The adgf mutant has lower ammonia levels (Fig. 6), diminished acaA expression (Fig. 7) and reduced cAMP levels (Fig. 7) in comparison to WT at both 12 and 16 h of development. Since ammonia is known to increase cAMP levels (Riley and Barclay, 1990; Feit et al., 2001), addition of ammonia addition to the mutant is likely to increase acaA expression, thereby rescuing the defects in cAMP signalling.
(3) No direct measurements in the dhkD mutant to show that it acts upstream of adgf in the control of changes in cAMP signalling and tip formation.
The histidine kinases dhkD and dhkC are reported to modulate phosphodiesterase RegA activity, thereby maintaining cAMP levels (Singleton et al., 1998; Singleton and Xiong, 2013). By activating RegA, dhkD ensures proper cAMP distribution within the mound, which is essential for the patterning of prestalk and prespore cells, as well as for tip formation (Singleton and Xiong, 2013). Therefore, ammonia exposure to dhkD mutants is likely to regulate cAMP signalling and thereby tip formation. We will address this issue by measuring cAMP levels in the dhkD mutant.
