Journal of Biotechnology, 22 (1992) 329-352

329

© 1992 Elsevier Science Publishers B.V. All rights reserved 0168-1656/92/$05.00

BIOTEC 00710

The decisive role of the Saccharomyces cerevisiae cell cycle behaviour for dynamic growth characterization Thomas Miinch, Bernhard Sonnleitner and Armin Fiechter Department for Biotechnology, Swiss Federal Institute of Technology, ETH Ziirich H6nggerberg, Ziirich, Switzerland (Received 2 July 1991; revision accepted 20 August 1991)

Summary The dynamic behaviour of the cell cycle and the physiology of Saccharomyces cerevisiae was monitored in transient experiments. Frequent flow cytometric analyses of the DNA (nuclear phase state) and the cell size enabled us to characterize the proliferation properties of yeast cells under well controlled and undisturbed cultivation conditions. Preliminarily, the correlation between flow cytometric light scattering measurements and the cell size was attested for yeasts. These flow cytometric results are compared with the physiological behaviour of the culture that was detected by high resolution on-line analyses and off-line measurements. The presented results focus on the importance of the yeast cell cycle behaviour for the dynamic growth characterization. Any kind of transients in yeast cultures induced partial synchronization. The characteristics and the time course of the yeast cell cycle were found to be strongly dependent on the physiological environment. Yeast cell cycle; Saccharomyces cerevisiae; Flow cytometry; DNA and cell size measurement; Dynamic experiment; High resolution analysis

Correspondence to: T. Miinch, Dept. for Biotechnology, Swiss Federal Institute of Technology, ETH Ziirich H6nggerberg, CH-8093 Ziirich, Switzerland.

330 Introduction

The segregated and structured understanding of "growth" of a cell population is of significant interest in biotechnology. Growth characteristics of many eukaryotic and prokaryotic cell types are described in the literature. However, reliable information about culture dynamics in the short time scale and under transient conditions is still missing. Part of these dynamics are assumed to be closely related to cell proliferation properties. Therefore, a better biological understanding of growth of a cell population needs the knowledge about both the physiology and cell cycle events. The cell cycle is doubtless the most decisive and central element during the life and growth of a cell. Different steps during the cell proliferation are characterized by typical, dramatically changing intracellular, (macro-) molecular compositions and cell shapes. Also environmental and cellular interactions are changing during the cell cycle. Much effort using biochemical and molecular biological methods has permitted the identification of key positions in the complex mesh of events during the cell cycle (Elliot and McLaughlin, 1983; Hartwell and Unger, 1977; Nurse and Thuriaux, 1984). The two proteins MPF (maturation promoting factor) and cyclin were identified as obvious intracellular control factors for the cell cycle initiation. Also a number of so called cdc gene products were found to influence the start or the velocity of different proliferation steps. However, the most obviously significant influence of extracellular signals like transmitter substances or growth substrate on the cell cycle mechanism has not yet been explained. Also the decisive factor of the cell cycle time course is not detectable by this kind of investigation. It is just this knowledge about the interactions between cell cycle events and external signals that is basic for the comprehension of the growth behavioor of a population. The controlled cultivation of the eukaryote Saccharomyces cerevisiae was considered to be a nearly ideal experimental setup for this kind of cell cycle study. The substrate offered to the cells was chosen as a controllable, external signal for cell cycle activity. With the high performance bioreactor equipment used (Sonnleitner and Fiechter, 1988; Sonnleitner et al., 1991) the substrate flux to the culture could be manipulated easily and with high reliability by the experimenter. Resulting responses of the cell cycle activity on changes of the quality and the quantity of the offered substrate could be detected by flow cytometry. Using this convenient method, two cell cycle related properties were measured in single cells: the relative DNA content to distinguish cells in the G1, S and G 2 / M phases of the cell cycle (Elliot and McLaughlin, 1983) and the relative cell size. The evidence of the advantages and possibilities of this experimental setup for cell cycle studies becomes obvious: (1) performance of cell cycle studies in well controlled and reproducible experiments; (2) defined environmental conditions (physical, chemical); (3) simultaneous observation of both, physiological and proliferation properties of a culture; (4) no disturbance of cellular interactions; (5) spontaneous synchronization mechanism of S. cerevisiae; (6) time is a directly measurable proliferation characteristic; (7) single cell measurements of DNA and

331 cell size are statistically significant: 20,000 cells analyzed per flow cytometry sample. The described approach in the investigation of the yeast cell cycle mechanisms and dynamics is fairly new compared with other methods. It was made possible first by the development of suitable bioreactor equipments. Genetic and biochemical methods are much more established in this area: these approaches have led already to common findings, to "the union of two views of the cell cycle" (Murray and Kirschner, 1989). New insight in cell cycle control by the approach described in this article, however, is still rather difficult to compare with genetic and biochemical knowledge because of considerably different perspectives. Some time will be necessary before one is able to see also the union with this view of the cell cycle. The application of flow cytometric methods for investigating the yeast cell cycle is quite established. Block et al. (1990) used slit scanning flow cytometry to analyze cell cycle-dependent events (cell size distributions) from cell populations grown in shake flasks. Inadequate cultivation techniques here did not allow the characterization of the actual growth state. In a similar way also, double staining techniques were used to determine D N A and a yeast antigen (Eitzmann et al., 1989). Scheper et al. (1987) characterized controlled yeast cultures by measuring DNA and cell size distributions in batch and continuous cultures. However, the low sampling frequency does not lead to a valid representation of the proliferation behaviour. Porro et al. (1988) showed total protein and cell volume distributions in synchronous yeast cultures and tried to project its periodic modifications to cell cycle events. This article shows for the first time results of a combined use of flow cytometric methods and well controlled cultures for the characterization of the S. cerevisiae cell cycle dynamics under transient cultivation conditions.

Materials and Methods

Organism The experiments were carried out with Saccharomyces cerevisiae, strain H 1022 (ATCC 32167).

Medium, cultivation conditions, bioreactor The synthetic cultivation medium D (Hug et al., 1974) containing 3% glucose was used. A mechanical foam separator replaced the use of antifoam agents. Sterilization occurred by microfiltration (0.2/zm, Millipore, U.S.A.). The following cultivation conditions were kept constant in the bioreactor, unless otherwise mentioned: temperature 30.0 + 0.05 o C, pH 5.00 + 0.02, gas flow rate 2.00 + 0.05 vvm, pressure 1.02 + 0.005 bar. A gravimetric medium flow measurement with a simple PID control loop led to an accuracy of < 5% of the flow rate setpoint. All cultivations were performed in a high performance bioreactor (Sonnleitner and

332 Fiechter, 1988). Sensors for pH, redox, fluorescence, p C O 2 and 0 2 (Ingold, Urdorf, Switzerland) were used routinely.

Automatic bioprocess control and data handling For the control of the bioprocess and the data acquisition an equipment following the concept of Sonnleitner et al. (1991) and Locher et al. (1991) was used. A direct digital slave computer (Alert 50, Alfa Laval) provided the link between hardware signals and software variables. On a secondary level a supervisory computer (VAX station) allowed interfacing with the process, collection, display and mathematical treatment of the data.

Staining procedures and flow cytometry DNA staining Cells were fixed in 70% precooled ethanol after harvesting, permeabilized with pepsin/HC1 (5.5 ml 1 M HC1, 0.5 g Pepsin [Merck no. 7190], 94.5 ml water) for 30 min at room temperature after washing in TRIS buffer (6.1 g TRIS [Merck no. 8382], 42 ml 1 M HC1, 958 ml water, titration to pH 7.5 with NaOH), then treated with RNAse for 1 h at 37 ° C (1 g I - l RNAse [Serva no. 34390] in TRIS buffer). After a second incubation with pepsin/HCl, ceils were stained over 1 h with propidium iodide (50 mg PI [Sigma no. P-4170], 14.2 g MgC12 • 6 H 2 0 , 21.8 g TRIS, 10.5 g NaC1 in 1000 ml water, pH 7) or DAPI (0.1 mg DAPI [Sigma no. D-1388], 8.13 g MgCl 2 • 6 H 2 0 in 1000 ml TRIS buffer). Then, the dye solution was replaced by TRIS buffer and the cells were stored at 4 ° C in darkness before the flow cytometric measurements (final cell concentration: approx. 106 ml-1). Bud scar staining A 250 /~l fresh culture sample was washed with 5 ml 0.9% aqueous NaC1 solution, then stained 15 min at room temperature with Calcofluor White M2R (2 mg 1- ~ fluorescent brightener 28, Sigma no. F-6259, in 0.9% NaC1). After a second washing step, the cells were resuspended in NaC1 solution to a final cell concentration of 106 ml-~ and finally stored not longer than 1 day at 4 °C in darkness. The flow cytometric measurements were performed on a Cytofluorograph 50 H coupled to the 2151 computer system (Ortho Diagnostics, MA, U.S.A.). Flow cytometry For measurements with DAPI and Calcofluor White M2R a C O H E R E N T I N N O V A argon-ion laser (model 300, 100 roW) was used operating at 351 nm. The separation of the forward scattered and the blue fluorescent light occurred with a 400 nm dichroic long-pass filter, the DAPI fluorescence was detected after band-pass filtering at 475 nm. Propidium iodide stained cells were illuminated with a C O H E R E N T I N N O V A argon-ion laser (model 70, 100 mW) operating at 488 nm, the separation of scattered and fluorescent light occurred with a 500 nm dichroic long-pass filter. The PI fluorescence was detected after passing a 590 nm long-pass filter. The Ortho histogram data files were analyzed or converted with

333 the MULTI-CYCLE program (Phoenix flow systems, San Diego, CA; order of the curve to fit the S-phase: 0) and the resulting data files further evaluated with MATLAB.

Other analyses The exhaust gas of the bioreactor was analyzed with a paramagnetic 0 2 analyzer (Servomex oxygen analyzer 540 A, Sybron Taylor, U.K.) and an infrared CO z analyzer (Binos 1, Leybold Heraeus, F.R.G.). Biomass dry weight concentrations were determined gravimetrically after membrane filtration (0.2 tzm, Zetapor). High frequency sample collection: for a reproducible and frequent sampling of cell free culture liquid a self constructed cross flow filtration unit was used. Residence time of the culture in the filtration loop: < 5 s. Minimalized dead volume of permeate behind the membrane filter (PVDF 0.2/.~m, MF 6502, LIGACON): 250 ~1. Permeate flux: > 1 ml min -1. Ethanol was measured by a gas chromatograph (5830 A, Hewlett Packard) at 140 °C using Porapack Q (Supelco Inc.) in a 6 ft stainless steel column.

Forward scatter and DNA measurements with Saccharomyces cerevisiae

Flow cytometric DNA measurements of S. cerevisiae cells have quite different properties compared with those of higher eukaryotes. The small amount of DNA per cell and the significant occurrence of extranuclear DNA in yeast cells lead to characteristic DNA histograms with broad population distributions. The correlation of the forward scatter intensity with the relative cell size is used in the following discussions. Because this correlation is by far not a general law, it needs to be verified. The most important properties of yeast DNA histograms and the evidence of relative yeast cell size measurements are therefore discussed briefly in the following two sections and should help to understand the interpretation of the results discussed later.

Yeast DNA histograms Flow cytometric DNA measurements in growing cells have typically the following properties: three more or less distinguishable subpopulations occur in DNA histograms, containing cells in the G1, the S phase and in the G2 or M phase of the cell cycle. Ceils with one full genome belong to the G1 phase of the cell cycle, cells with the double amount of nuclear DNA are entering into the G2 phase after completed DNA doubling and then start the mitosis phase (M phase). Cells in the G2 and the M phases cannot usually be differentiated by flow cytometry because of their equal DNA content. The S phase is the period of nuclear DNA synthesis. It should further be mentioned that cells that have completed mitosis, but have not yet separated, belong to the G1 phase. In the DNA histograms these double cells are particles with two full genomes and are therefore a part of the subpopulation called G 2 / M in the following. From the biological point of view, cells in the

334

800 ~

'G1

700~

....

..............

600

~ ~°°r ~°°U ~ ~°°V -~

rv~-[]

i . . . . . . . .

...... i. . . . . .

', G 2 + M

...............................................................

.........dillltiill ...........li]ilLtiNlh..................... 1. . . .lllillllf . .............. tllilll]ll[~ ~ ........

...................

....

~oo[

~ 1 I ] 1 1 I ~

0

I..............

'~

10

z0

ao relative

........

.........

............ .......

....

.................. ~ 1 1......1 1 1 ........ I ~

40 DAPI

50

60

70

.........

o0

90

100

fluorescence

Fig. 1. A typical flow c y t o m e t r i c yeast D N A m e a s u r e m e n t , s t a i n e d with D A P I . The relative f r e q u e n c y of cells is p l o t t e d a g a i n s t the c o r r e s p o n d i n g ( D A P I - ) f l u o r e s c e n c e class n u m b e r s from 1 to 100. T h e cells w e r e t a k e n out from a s y n c h r o n o u s l y growing culture. T h e q u o t i e n t of the G1 and the G 2 + M m e a n f l u o r e s c e n c e values is 2.1.

G 2 / M phase must also have the double nuclear DNA content compared with cells in the G1 phase. This implies, for flow cytometry measurements, that after a reliable and specific DNA staining procedure the average amount of cells in the G 2 / M phase have to occur at exactly the double fluorescence value of the fluorescence mean value of the G1 cells. For most of the eukaryotic cells (e.g. mammalian cell lines) this behaviour can be observed (Shapiro, 1988). DNA histograms of yeast cells, however, show quotients of the mean values of the G1 and the G 2 / M fluorescence between 2.1 and 2.3 (Fig. 1) in spite of a specific DNA staining procedure (DAPI binds specifically to DNA). In fact, the cell cycle dependent, variable amount of extranuclear DNA is quantitatively not negligible in yeast cells: Hartwell (1974) described that 5 to 20% of the total cellular DNA in yeast cells is mitochondrial DNA. Fluorescence microscopic observations and fluorescence scanning measurements of DNA stained cells showed remarkable variations of the mitochondrial D N A content depending on the cell cycle phase. Yeast cell size measurements

The comparison of the forward scattered light (small-angle scatter) with the cell size in flow cytometric measurements has been widely used to estimate the cell size since it was demonstrated by Mullaney et al. (1969) that the intensity of light scattered at small angles (0.5-2 ° ) from an incident laser beam was proportional to

335 100

~

The decisive role of the Saccharomyces cerevisiae cell cycle behaviour for dynamic growth characterization.

The dynamic behaviour of the cell cycle and the physiology of Saccharomyces cerevisiae was monitored in transient experiments. Frequent flow cytometri...
1MB Sizes 0 Downloads 0 Views