Changes in Cell Size and DNA Content in Sulfolobus Cultures during Dilution and Temperature Shift Experiments

Strains.

S. acidocaldarius (DSM 639) and S. solfataricus (DSM 1616) type strains were purchased from the Deutsche Sammlung von Mikroorganismen und Zellkulturen in Braunschweig, Germany. S. shibatae B12 was a kind gift from Christiane Elie.

Culture media and growth conditions.

The growth medium used for S. acidocaldarius and S. solfataricus was Allen mineral base supplemented with 0.2% tryptone (5). S. shibatae was grown in the medium described by López-García and Forterre (11), except that Ca(NO₃)₂ was omitted. Solid medium (7) was obtained by addition of 0.6% Gelrite (Merck), 10 mM MgSO₄, and 2.5 mM CaCl₂ (final concentrations). The cultures were incubated in Erlenmeyer flasks at 79°C in water baths at a shaking speed of 200 rpm, and growth was monitored by optical density (OD) measurements at 600 nm. To obtain stationary-phase cells for the dilution experiments, the cultures were incubated for a further 10 to 20 h after the OD had ceased to increase. In the dilutions, the flasks and media were always first preheated at 79°C. In the temperature shift experiments, liquid cultures were inoculated such that the strains had been growing exponentially for at least 10 generations before the shifts, which were performed at an OD of ≤0.2.

Sampling, preparation of cells, combined phase and epifluorescence microscopy, and flow cytometry.

Sampling and preparation of cells, as well as phase-fluorescence microscopy and flow cytometry, were performed as described previously (4, 13). In the flow cytometry, cell size was measured as light scatter. We confirmed that the gradual increase in light scatter observed in the dilution experiment indeed reflected increased cellular size, by direct microscopy measurements (data not shown).

Dilution of stationary-phase cultures.

The strains were grown in liquid cultures at 79°C (see Materials and Methods for strains and culture media). The cultures were considered to have reached stationary phase when the OD had ceased to increase and the cells had undergone the morphological changes described in the Introduction. They were then diluted 5- to 10-fold into fresh growth medium and sampled for phase-fluorescence microscopy and flow cytometry at different time points after the dilution.

Before the dilution, the cells were small and contained 2 genome equivalents (Fig. 2B, first row). As observed previously (4), a minor part of the population appeared to contain 3 chromosome equivalents in stationary phase.

An immediate increase in light scatter was evident upon dilution of the culture (Fig. 2B, second row). As only 60 s had passed since the previous measurement, cellular growth could not account for this increase which, instead, may represent osmotic effects on cell size and shape when fresh medium is added.

During the following 240 min, there was a gradual increase in light scatter (Fig. 2B, first column), showing that cellular growth occurred and that this growth started rapidly after dilution. The early growth response was also evident in OD measurements (Fig. 2A). The DNA content of all cells remained at the predilution level (Fig. 2B, second column), showing that no chromosome replication or cell division occurred during this long period of continuous cellular mass increase.

The first sign of cell division, detected at 270 min, was evident as a small peak of cells containing a single chromosome (partly obscured by peaks from later time points in Fig. 2B but clearly visible in Fig. 2C). This peak increased significantly in relative size over the following time points, showing that a substantial part of the cell population divided. In parallel, the average cell size ceased to increase and, instead, decreased somewhat (evident as a decrease in the right part of the light scatter peak, corresponding to large cells), as expected if cell division was occurring. The low resolution between the 1- and 2-chromosome peaks at 360 min indicated that a substantial fraction of the cell population had initiated replication well before this time point and therefore contained between 1 and 2 chromosome equivalents. By 480 min, the typical cell size and DNA content distributions of an exponentially growing cell population had become established (Fig. 2B, bottom row).

A higher-resolution analysis at around the time points at which cell division and chromosome replication were initiated is shown in Fig. 2C. Cell division was first initiated at around 270 min after the dilution or slightly less than 1 doubling time (288 min) under these growth conditions. Although it is difficult to pinpoint exactly, we estimate from Fig. 2C that the start of chromosome replication occurred no later than 30 to 40 min after division, at around the 300- to 310-min time points. This is probably an overestimate since the replicating part of the population must be relatively large before it starts to affect the DNA distributions.

The cells must carry out nucleoid segregation in advance of cell division to ensure that each daughter cell receives a chromosome. To estimate the time point at which this occurred, we determined the proportion of the cell population that displayed segregated nucleoids at different time points (Fig. 3). A low level of cells with two fluorescence foci was already present at the start of the experiment. The significance of this background population is unknown. The proportion of cells with segregated nucleoids started to increase about 3 h after the dilution and reached a maximum of about 15% at around 310 min. The proportion of the population that displayed visible cell constriction increased from 0 to about 1% at around the 260-min time point and reached a maximum of approximately 5% at 340 min.

The results were essentially the same regardless of whether the cultures had been kept for 10 or 20 h in stationary phase before the dilution (data not shown).

Dilution of exponentially growing cultures.

Dilution experiments were also performed with exponentially growing cultures to investigate the effects on chromosome replication and cell division and to determine whether balanced growth was maintained.

Exponentially growing cultures were diluted 2- to 10-fold at an OD of about 0.1. Apart from a slight decrease in light scatter, no immediate effects were observed in the flow cytometry analysis (Fig. 4A, time points between −1 and 50 min). Unexpectedly, the proportion of cells in the prereplication (peak with cells containing a single chromosome) and replication (cells with a DNA content between 1 and 2 chromosome equivalents) cell cycle periods decreased significantly after the 50-min time point and had largely disappeared by 80 min. Furthermore, the OD curve flattened out at about 30 min after the dilution (Fig. 4B), showing that culture growth decreased significantly.

After approximately an additional 60 min, the OD started to increase exponentially again. At the same time point, around 90 to 100 min after the dilution, flow cytometry (Fig. 4A) revealed that division and replication restarted in the cell population and the characteristic exponential DNA distribution was rapidly re-established. The light scatter parameter remained essentially unaffected throughout the experiment, showing that the cells did not increase significantly in size during the period of growth decrease and division inhibition.

Transient temperature shift experiments.

Members of the genus Sulfolobus are hyperthermophiles that grow optimally at temperatures of around 80°C. We wanted to investigate the effects of shifts to lower temperatures on growth and cell cycle characteristics. Flasks containing exponentially growing cultures were therefore removed from a shaking water bath and placed in other shaking water baths either at room temperature or in ice-water. After different time periods (between 10 min and 8 h), the flasks were returned to the 79°C water bath. The OD was monitored throughout the shifts, and samples for flow cytometry were removed at regular time intervals.

A control culture kept at 79°C continued to grow until it eventually reached stationary phase at an OD of approximately 0.8 (Fig. 5A). The OD of the control culture increased little during the first hour after parts of the culture had been removed, possibly due to changes in aeration or other parameters as a consequence of the decrease in both culture volume (from 80 to 15 ml) and flask volume (from 400 to 100 ml).

A shift to room temperature resulted in immediate cessation of the OD increase (Fig. 5A), similar to the control culture. However, in contrast to the transient effect in the control, the OD remained constant during the entire time at low temperature. In flow cytometry, no effects on either cell size or DNA content distribution were detected during the corresponding time interval (Fig. 5B and C).

Upon a return to 79°C, the OD increase resumed after a short lag period, regardless of whether the culture had been incubated for 5 or 8 h at low temperature (Fig. 5A). Remarkably, almost no effects on cell size or DNA content distribution were apparent (Fig. 5B and C), as if the cultures simply continued to grow with their preshift characteristics without the cells “noticing” the dramatic temperature downshift. In the samples from the overnight cultures (which had not yet reached stationary phase), the relative size of the peak corresponding to <1 chromosome equivalent in the DNA distributions increased. It is possible that after a time delay, DNA degradation occurred in part of the cell populations.

In cultures shifted to ice and back to high temperature, the effects of the shifts (not shown) were more dramatic. As in the cultures shifted to room temperature, no effects were detected by flow cytometry during the incubation on ice. However, a large drop in colony-forming efficiency on solid medium occurred, giving the impression that a large part of the cell population rapidly died on ice. However, the effect was reversible such that a rapid increase in plating efficiency took place after a return to 79°C and the drop in plating efficiency therefore instead indicates that the cells became hypersensitive to dilution and/or plating on solid medium after transfer to ice. The growth lag phase after a return to 79°C was generally longer for the cultures shifted to ice than for those shifted from room temperature, and the growth resumption kinetics varied between experiments. Importantly, the proportion of DNA-less cells increased dramatically over time after a return to high temperature, indicating that extensive DNA degradation occurred, in contrast to the more limited degradation observed in the cultures shifted to room temperature as described above.

In contrast to the experiments in which exponentially growing cultures were diluted into fresh medium as described above, none of the temperature-shifted populations went through a period when the proportion of cells in the prereplication and replication periods transiently decreased.

Other Sulfolobus species.

Three different Sulfolobus species, S. acidocaldarius, S. solfataricus, and S. shibatae, were studied, but due to space limitations, only the S. acidocaldarius results are illustrated. When the experiments were performed with S. solfataricus and S. shibatae, the results were similar in most respects. The only exception was that there was a larger variation in the time period from the dilution event until the first signs of cell division (usually between 4 and 6 h, occasionally longer), particularly for S. solfataricus, whereas for S. acidocaldarius the time interval of 4.5 h showed less variation between experiments. Nucleoid segregation and cell constriction times were not determined for S. solfataricus and S. shibatae.