Plasticity and the evolution of group-level regulation of cellular differentiation in the volvocine algae

During the evolution of multicellularity, the unit of selection transitions from single cells to integrated multicellular cell groups, necessitating the evolution of group-level traits such as somatic differentiation. However, the processes involved in this change in units of selection are poorly understood. We propose that the evolution of soma in the volvocine algae included an intermediate step involving the plastic development of somatic-like cells. We show that Eudorina elegans, a multicellular volvocine algae species previously thought to be undifferentiated, can develop somatic-like cells following environmental stress (i.e., cold shock). These cells resemble obligate soma in closely related species. We find that somatic-like cells can differentiate directly from cold-shocked cells. This differentiation is a cell-level trait, and the differentiated colony phenotype is a cross-level byproduct of cell-level processes. The offspring of cold-shocked colonies also develop somatic-like cells. Since these cells were not directly exposed to the stressor, their differentiation was regulated during group development. Consequently, they are a true group-level trait and not a byproduct of cell-level traits. We argue that group-level traits such as obligate somatic differentiation can originate through plasticity and that cross-level byproducts may be an intermediate step in the evolution of group-level traits.

Methods
Induction of somatic-like cells over multiple generations

In order to determine whether somatic-like cells develop in response to cold shock, we inoculated a 250 mL flask of Standard Volvox Media (SVM) with E. elegans UTEX 1201 and grew the cultures on a shaker at 25°C with a 16:8 hour light:dark cycle. After 5 days, we transferred 0.5 mL of the culture to a new 250 mL flask with fresh SVM and allowed the culture to grow for 10 days. We then placed 1 mL of the culture with young unexpanded colonies in six Eppendorf tubes. Three of the tubes were cold-shocked in a covered ice bath at 1° C for two hours while the other three control tubes were placed in the dark at 25° C for two hours. The contents of each tube were then transferred to fresh media in separate wells of a 6 well plate.

 To assess the development of somatic cells over multiple generations following cold shock, we imaged the cold-shocked colonies (which we refer to as generation 1; G1), their offspring (generation 2; G2), and their grand-offspring (generation 3; G3) (Figure 1F). We did the same for three generations of controls. Wells were examined daily for nine days to determine whether colonies were reproducing and to track their offspring. The density of each well was measured at the time of imaging by mixing the culture and then counting the number of colonies contained within 50 µL. During this experiment, we recorded the number of somatic-like cells each colony had. We imaged approximately 100 expanded adults (Figure 1E and Figure 2) from the cold treatment and 100 expanded adult controls from each generation. Each well was imaged for a single generation and treatment.

When imaging, we mixed the content of each well before randomly withdrawing colonies and placing them on a grid slid. We then systematically imaged all expanded adults on the slide using a Nikon Eclipse Ti-E inverted microscope. We used both brightfield and fluorescence microscopy to image colonies. For each expanded adult colony, we characterized each cell by size and by the presence/absence of chlorophyll autofluorescence. Chlorophyll autofluorescence was used as a proxy to determine whether the cell was likely to be alive. The basic experimental setup is visually summarized in Supplementary Material S1.

Assessment of the heritability of somatic-like cells 

In order to specifically determine when somatic-like cells develop following cold shock and whether the phenotype is heritable, we grew two flasks on a shaker in the conditions described in the previous section. From these cultures, we inoculated two new flasks at a density of 5 colonies per mL and measured the culture density every other day. After 10 days, 48 colonies at the same developmental stage (unexpanded juveniles; as represented in Figure 1E and Figure 2) were haphazardly chosen from the two flasks and imaged to confirm they did not have visible somatic-like cells. Each of the 48 colonies were placed in separate Eppendorf tubes with 40 µL of SVM. We placed 24 of the Eppendorf tubes in an enclosed, dark ice bath at 1°C for two hours and the other 24 in a dark container at 25°C. After two hours, the contents of each Eppendorf tube were transferred into fresh media in separate wells of two 24 well plates. We imaged all G1 colonies before they divided and gave rise to G2 offspring colonies. After they hatched, we transferred individual G2 colonies to separate wells in 24 well plates and imaged them. We imaged G3 colonies from the wells they hatched out in. Imaging was conducted with a Nikon Eclipse Ti-E (Nikon, Tokyo, Japan). A summary of the three generations is shown in Figure 1F and Figure 2.

Our final sample sizes were affected by colony survival, successful transfer to or from wells, and the number of reproductive cells (and therefore number of offspring) present in generations 1 and 2. We collected data from 12 cold-shocked G1 colonies, 19 control G1 colonies, 72 cold-shocked G2 colonies, 199 control G2 colonies, 661 cold-shocked G3 colonies, and 1267 control G3 colonies.

Each time we imaged a colony, we recorded the number of large (reproductive) and small (somatic-like) cells. Since we tracked the differentiation state of each colony and that of their offspring, we were able to characterize the heritability of the facultatively differentiated phenotype. Specifically, we determined if facultatively differentiated colonies had a higher proportion of differentiated offspring than undifferentiated colonies. The data reported here are the result of two independent experiments completed several months apart.

Separation of single cells

To determine whether somatic-like cells can develop in the absence of interactions among cells, we grew colonies following the protocol described above. We separated two single cells from 20 unexpanded undifferentiated colonies using needles to break apart the colonies and placed each cell into a separate Eppendorf tube, along with 40 µL of SVM. Twenty tubes, each containing 1 cell from the 20 colonies, were placed in an ice bath for 2 hours. The other 20 tubes were placed in the dark at room temperature for 2 hours and used as controls. Each of the 40 cells was placed on a separate 1% agar plate (in SVM) to prevent cell-cell communication and ensure that cells could be clearly seen and tracked. Four to six days after the treatment, each plate was imaged using a Nikon Eclipse Ti-E inverted microscope. The experiment was replicated with an additional 40 cells from 20 unexpanded colonies.  

Cell size measurements

To determine if somatic-like cells are significantly smaller than reproductive cells and the same size as somatic cells in other species, we measured reproductive and somatic-like cells in mature E. elegans UTEX 1201 colonies as well as reproductive and somatic cells in Pleodorina starrii NIES 1362 and the somatic cells in Volvox carteri Eve. We inoculated two 250 mL flasks with P. starrii and V. carteri on a shaker for five days in the conditions described above. We repeatedly withdrew 10 µL of media from the flask and imaged the expanded adult colonies on a slide using a Nikon Eclipse Ti-E inverted microscope (Nikon, Tokyo, Japan).

To measure the sizes of somatic-like cells and reproductive cells from E. elegans UTEX 1201 expanded adult G1 and G2 colonies, we followed the same protocol described in the sub-section “Induction of somatic-like cells over multiple generations.” We then measured cell sizes using ImageJ (31). We traced the perimeter of each cell in a colony using the circle tool in ImageJ, calibrated our measurements using the scale bar from each image, and obtained the minimum diameter, maximum diameter, and area of each cell.

Assessment of reproductive potential

Somatic cells in obligately differentiated species are terminally differentiated and do not give rise to the next generation. To determine if somatic-like cells reproduce, we set up experimental well plates using the protocol described above. After the offspring of cold-shocked G1 colonies hatched, the G1 shells composed of extracellular matrix and somatic-like cells were transferred to new wells. The wells that contained shells were monitored for five days to determine if any of the somatic-like cells went on to reproduce. The same procedure was followed for the G2 colonies. To address the possibility that cold treatment and culture conditions could inhibit reproduction, twelve differentiated cold-shocked colonies were transferred to separate wells and after reproduction, their offspring remained in the wells. Additionally, 12 control colonies were transferred to separate wells and their offspring remained in the wells after reproduction. We examined these wells after 5 days and recorded whether viable colonies were still present to ensure that the growth conditions didn’t inhibit reproduction.