Protein turnover in relation to maintenance metabolism at low photon flux in two marine microalgae


  • A. QUIGG,

    Corresponding author
    1. School of Biological Sciences, Monash University, Clayton, Victoria, 3800, Australia
      Correspondence: A. Quigg. Fax: +1 732 932 4083; e-mail:
    Search for more papers by this author
    • *

      Present address: Environmental Biophysics and Molecular Ecology Program, Institute of Marine and Coastal Sciences, Rutgers, The State University of New Jersey, 71 Dudley Road, New Brunswick, New Jersey, 08901, USA.


    1. School of Biological Sciences, Monash University, Clayton, Victoria, 3800, Australia
    Search for more papers by this author

Correspondence: A. Quigg. Fax: +1 732 932 4083; e-mail:


Acclimation to very low photon fluxes involves adjusting a suite of physiological characteristics that collectively elicit a physiological response. Facilitating such changes is pro-tein turnover. Dunaliella tertiolecta (Butcher) and Phaeodactylum tricornutum (Bohlin) were grown in turbidostats at a range of photon fluxes between 2 and 300 µmol photons m−2 s−1. The kinetics of pulse-chase labelling of the protein with 3H showed that (1) two protein pools were present, one of which turned-over rapidly (hours), and a second which turned over more slowly (days); and (2) protein turnover rates were slower in P. tricornutum than in D. tertiolecta. Phaeodactylum tricornutum had a lower maintenance coefficient for protein turnover than D. tertiolecta, and correspondingly a smaller proportion of its respiratory demands (30%) were associated with protein turnover than in D. tertiolecta (36%). There appears to be a correlation between lower metabolic activity, requiring lower protein concentrations, and an associated decreased cost of maintenance processes in P. tricornutum compared to D. tertiolecta. Differences between protein turnover rates and maintenance metabolic costs may be one of the photo-acclimation strategies that determine which photon niches microalgae can successfully exploit.


Major limitations on photolithotrophic growth at extremely low photon fluxes are thought to be protein turnover, charge recombination in photosystem II, and proton leakage and slippage (Raven, Külber & Beardall 2000). These processes act synergistically, limiting rates of photosynthesis, and affecting the minimum specific growth rate that eukaryotic microalgae can sustain. Protein turnover will be defined as a cycle of degradation and re-synthesis, and is the focus of the present investigation. This essential process allows cells to re-utilize amino acids, to exchange protein content during growth, and to acclimate to environmental stimuli (Huffaker & Peterson 1974; Davies 1982; Vierstra 1993). Protein turnover is believed one of the most significant components of maintenance processes (and hence dark respiration) in terms of total cell energy requirements (Penning de Vries 1975; Raven & Beardall 1981; Geider & Osborne 1989; De Visser, Spitters & Bouma 1992; Zagdañska 1995; Geider, MacIntyre & Kana 1996; Lambers, Chapin & Pons 1998; Raven et al. 2000). This has important implications for plant carbon balance, as any factor that increases protein concentration will significantly increase maintenance costs. The amount of quantitative data on protein turnover as a function of maintenance costs for higher plants is increasing (Van der Werf et al. 1992; De Visser et al. 1992; Bouma et al. 1994; Zagdañska 1995; Scheurwater et al. 2000). However, for algae, the reports are scarce (Richards & Thurston 1980a,b; Thurston & Richards 1980).

Growth of marine microalgae is likely to be frequently light-limited, particularly during intense mixing of the euphotoic zone and in the deep-water chlorophyll maximum layer. The diatom Phaeodactylum tricornutum has been shown to grow at ≤ 1 µmol photons m−2 s−1 (Geider, Osborne & Raven 1985, 1986). Dunaliella tertiolecta, a chlorophyte, exhibits 20-fold higher compensation photon fluxes for growth (Icg) (Falkowski & Owens 1980; Quigg 2000). The parameter Icg defines the point at which photosynthesis and respiration rates are in balance, so that the net growth rate is zero. One of the major differences between high- and low-Icg microalgae is thought to be their maintenance costs (Geider et al. 1985). Maintenance respiration rates can be used to estimate (assuming complete coupling in respiratory energy conversions) the energy available for maintenance processes. The likelihood that some of the maintenance energy requirements in the light are supplied by the direct use of photoproduced adenosine triphosphate (ATP) rather than via respiration cannot be ignored, but is difficult to quantify and would have little effect on the overall energy budget (Raven & Beardall 1981). In addition, Bouma et al. (1994) suggested that many of Penning de Vries (1975) calculations need to be reviewed in the light of new information on the process(es) associated with protein (re)synthesis.

Given this, we have quantified the rates of protein turnover in P. tricornutum and D. tertiolecta using a 3H labelling and dilution (pulse-chase) method modified for use with microalgae. The energy demand of the measured rate of protein turnover was then compared with the energy available for maintenance processes under the same conditions. Using this information, the fraction of maintenance energy associated with protein turnover was estimated. Overall, maintenance costs for D. tertiolecta were significantly higher than those of P. tricornutum. These relation- ships were considered in relation to the different photo-acclimation strategies used by D. tertiolecta and P. tricornutum, and their effect on the minimum photon flux for photolithotrophic growth.


Growth conditions

Dunaliella tertiolecta (CS-175) and P. tricornutum (CS-29) were obtained from the CSIRO Culture Collection of Microalgae (CSIRO Division of Oceanography, Tasmania, Australia) and grown in 500 mL turbidostats in the artificial seawater medium, PHK (a modified ‘D’ medium; Provasoli, McLachlan & Droop 1957). The cultures were grown under 24 h continuous illumination at 18 °C for a minimum of 5–10 generations (depending on the growth rate) at seven photon fluxes between 2 and 300 µmol photons m−2 s−1.

For the protein turnover experiments, D. tertiolecta and P. tricornutum were grown at four photon fluxes corresponding to (1) a photon flux close to their Icg that will be designated µmin, where the minimum growth rate was measured; (2) two growth rates corresponding to 25 and 50% of the maximum growth rate; and (3) the maximum growth rate (µmax or 100%) (Table 1).

Table 1.  Cellular composition and photosynthetic characteristics of D. tertiolecta and P. tricornutum as a function of the photon flux for growth
OrganismPhoton flux for growth (µmol photons m−2 s−1)Growth rate (µ) as a function of µmaxaPmC (10−5 mol O2 mol C−1 s−1)C : N (mol C mol−1 N)Chl a : C (mg Chl a : mg−1 C)Protein pool size (pg protein cell−1)
  • a

    Refer to text for further details.

  • Values are averages (± SE). n.d, not measured.

D. tertiolecta
P. tricornutum 30µmin1.33 (0.08)5.16 (0.86)0.06915.8 (2.7)
 4525%2.20 (0.09)6.49 (0.16)0.066 8.75 (8.0)
 7050%2.94 (0.16)8.85 (0.05)0.084 8.24 (1.9)
100 6.34 (0.45)4.90 (0.14)0.103 
130µmax (100%)6.06 (0.55)3.01 (0.17)0.072 8.12 (2.2)
220 2.33 (0.36)4.32 (0.17)0.041 
285 1.66 (0.13)5.07 (0.33)0.033 
   2.5µmin2.14 (0.09)nd0.075 3.85 (0.4)
  625%2.18 (0.15)5.11 (0.11)0.046 2.73 (0.6)
 1550%2.29 (0.13)5.77 (0.18)0.045 3.49 (2.1)
 45 2.33 (0.24)5.14 (0.13)0.034 
 85 1.72 (0.07)5.38 (0.12)0.056 
100µmax (100%)1.83 (0.08)6.21 (0.08)0.021 2.34 (0.5)
140 3.04 (0.18)ndnd 

Protein turnover experiments were not carried out directly in turbidostats (due to constraints on equipment and a requirement to minimize the amount of 3H2O used), but in smaller (20 mL) chambers. These included an air inlet to aerate and mix cultures and an air outlet that was passed through to a desiccated silica gel trap. During chase experiments, cells were kept in the mid-exponential growth phase by periodic dilutions with fresh PHK medium. This ensured that cultures were neither nutrient-limited nor self-shading. Cell biomass was monitored at 24-hourly intervals by measuring in vivo chlorophyll (Chl) fluorescence with a Hitachi 2000 spectrofluorometer (Hitachi Australia Ltd, North Ryde, NSW, Australia), using excitation and emission wavelengths of 430 and 680 nm, respectively.

Gas exchange measurements

Samples drawn from turbidostats were used to prepare photosynthesis versus irradiance curves for D. tertiolecta and P. tricornutum. A Clark-type Hansatech DW1 oxygen electrode (Hausatech Instruments Ltd, Kings Lynn, Norfolk, UK) was used to measure rates of dark respiration (Rd), followed by rates of O2 evolution at 12 increasing photon fluxes between 2 and 2000 µmol photons m−2 s−1, at 18 °C. A slide projector fitted with a quartz halogen globe and appropriate neutral density filters was used for illumination. The photon flux incident on the internal surface of the electrode chamber was measured using a Li-Cor Integrating Quantum Radiometer/Photometer (Li 188B; LiCor Inc., Lincoln NE, USA). Cells were exposed to each photon flux for 3–5 min, by which time rates of O2 exchange had stabilized. Photosynthesis versus irradiance data were modelled using the equation of Platt, Gallegos & Harrison (1980), and normalized to cell number, cellular carbon or Chl a. Mean O2 exchange rates (± SE) were then estimated from a minimum of six replicate curves performed on cells grown at each photon flux (Table 1). A Superior Improved Neubauer haemocytometer (ProSciTech, Thuringowa, QLD, Australia) was used for cell counts. Cellular carbon and nitrogen were determined on the same cultures (n ≥ 3), dried onto Whatman GF/C filters (Whatman International Ltd, Maidstone, Kent, UK), and analysed with a Carlo ERBA CHN analyser (ThermoFinnigan, Hemel Hempstead, Herts, UK). Pigments were extracted with 90% acetone; Chl a was then determined using the equations of Jeffrey & Humphrey (1975).

The ratio of dark respiration rate to maximum photosynthetic rate (Rd : Pm) has been previously reported to be a sensitive indicator of the physiological state of algal cells (Verity 1982; Geider & Osborne 1989; Falkowski & Raven 1997). During protein turnover experiments, Rd and Pm were measured in cells growing in both turbidostats and chambers. Rd was measured first, followed by Pm once cells had been illuminated at 500 µmol photons m−2 s−1 for 7–10 min. Photosynthesis versus irradiance curves had established this photon flux as saturating for O2 evolution, but not photo-inhibitory.

Pulse-chase experiments

Protein turnover was measured by modifying the 3H2O labelling method of Humphrey & Davies (1975, 1976) for use with microalgae. A reverse isotope dilution in which 3H2O was incorporated into the protein pools (pulse experiment – 24 h) and an isotope dilution, in which labelled proteins were chased with endogenous/exogenous H2O (chase experiment) was performed. Dunaliella tertiolecta and P. tricornutum suspensions were divided into five chambers: three were inoculated with 18.5 kBq mL−1 3H2O and two were unlabelled (controls). Aliquots were removed hourly to monitor the time course of 3H incorporation into proteins and growth (Fig. 1). Both the protein content (see below) and proportion of labelled protein [via liquid scintillation counting with aqueous counting scintillant II (ACS II); Amersham Biosciences Pty, Ltd, Castle Hill, NSW, Australia] were measured. After the pulse, unincorporated label was removed from cells with three washes (10 mL) of fresh PHK medium (2500 × g, 10 min). Cells were re-suspended in fresh medium to start the chase experiment. Aliquots were again removed at regular intervals.

Figure 1. Time course of 3H.

Figure 1. Time course of 3H.

incorporation into proteins (right-axis) was monitored over a 24 h pulse-experiment. Cell counts (left-axis) were used to follow growth. This is a representative figure, taken from an experiment on D. tertiolecta grown at 45 µmol photons m−2 s−1 (25% µmax).

Although the data fitted a two-component first-order decay model well in many cases (r2 > 0.95), it was more realistic to regard the classes of protein in the samples as a heterogenous assembly where the observed decay rates were average values of a range of individual rates. The rate constant for protein degradation (kd; d−1) was the negative of the slope of a plot of the natural log of 3H-protein activity against time. The rate constants for D. tertiolecta and P. tricornutum were corrected for growth rate dilution by subtracting the growth rate (kd = rate constant − µ). Three independent pulse-chase experiments (as a minimum) were performed at each growth rate for both species.

Determination of total protein concentration

Cell samples were thawed on ice, vortexed, and then centrifuged (15 000 × g, 5 min). The supernatant was discarded, the pellet washed, and the cells re-suspended in fresh medium. For D. tertiolecta, this step was repeated to remove extracellular material. Phaeodactylum tricornutum cells were disrupted with a Branson sonicator (Branson Ultrasonic Corporation, Danbury, CT, USA) (2 × 15 s bursts), on ice, and centrifuged (15 000 × g, 5 min). Pellets were then re-suspended in ice-cold acetone (90%) and kept on ice for 30 min to extract pigments. Samples were then centrifuged (15 000 × g, 5 min), the supernatant discarded, and the pellets air-dried. These were then re-suspended in 500 µL 0.1 m NaOH/1% sodium dodecyl sulphate, incubated at 100 °C for 5 min, and then centrifuged (15 000 × g, 5 min). Protein in the supernatant was assayed at 750 nm according to Peterson (1977) using a bovine serum albumin standard curve as a reference.


Response of the various physiological parameters to the photon flux for growth

Growth rates of D. tertiolecta and P. tricornutum(Fig. 2a & b, respectively) exhibited a curvilinear response with photon flux for growth. Maximum growth rates were 1.4 and 1.15 d−1 for D. tertiolecta and P. tricornutum, respectively. Icg was calculated using the first three linear points on the growth curves, giving values of 18 µmol photons m−2 s−1 for D. tertiolecta and 2.4 µmol photons m−2 s−1 for P. tricornutum.

Figure 2. Growth rates of D. tertiolecta (a) and P. tricornutum (b) varied with photon flux between 0–300 µmol photons m−2 s−1.

Figure 2. Growth rates of D. tertiolecta (a) and P. tricornutum (b) varied with photon flux between 0–300 µmol photons m−2 s−1.

. Error bars (SE; n ≥ 8) in some cases were smaller than symbols.

Carbon-specific dark respiration rates (RdC) and maximum photosynthetic rates (PmC) increased linearly with growth rate in D. tertiolecta, up to approximately 200 µmol photons m−2 s−1(Fig. 3a; Table 1). RdC and PmC did not change as a function of the photon flux used to grow P. tricornutum (Fig. 3b; Table 1). The Chl a : C ratio (Table 1) and C and N quotas (Fig. 4a & b) decreased with the photon flux for growth in D. tertiolecta and P. tricornutum. Small increases in the C and N quotas, however, were seen at photon fluxes> 200 µmol photons m−2 s−1 for D. tertiolecta (Fig. 4a) and> 100 µmol photons m−2 s−1 for P. tricornutum (Fig. 4b). Despite variations in C and N concentrations, the average C : N ratio was 5.29 ± 1.76 mol C mol−1 N for D. tertiolecta and 5.53 ± 0.42 mol C mol−1 N for P. tricornutum (Table 1). Total protein concentration in P. tricornutum (3.10 ± 0.8 pg protein cell−1) did not change as a function of the photon flux for growth (Table 1). Only in D. tertiolecta growing close to their Icg (30 µmol photons m−2 s−1) was there a noticeable increase in the size of the protein pool (Table 1). At all other photon fluxes for growth, the protein pool of D. tertiolecta was some 53% smaller (8.37 pg protein cell−1). Differences in the physiological parameters (RdC, PmC, protein pool) of D. tertiolecta and P. tricornutum suggest these microalgae use different photo-acclimation strategies.

Figure 3. Specific dark respiration rates (RdC) increased with growth rate in D. tertiolecta (a). The linear regression describing this relationship is y = 1.0x + 0.065 (r2 = 0.66). There was no significant change in RdC with growth rate (P > 0.05) in P. tricornutum (b). Error bars (SE; n ≥ 6) in some cases are smaller than the symbols.

Figure 3. Specific dark respiration rates (RdC) increased with growth rate in D. tertiolecta (a). The linear regression describing this relationship is y = 1.0x + 0.065 (r2 = 0.66). There was no significant change in RdC with growth rate (P > 0.05) in P. tricornutum (b). Error bars (SE; n ≥ 6) in some cases are smaller than the symbols.

Figure 4.

Cellular C (filled diamonds) and N (open diamonds) quotas decreased with photon flux for growth in D. tertiolecta (a) and P. tricornutum (b). Small increases were seen at photon fluxes> 200 µmol photon m−2 s−1and 100 µmol photon m−2 s−1 for D. tertiolecta (a) and P. tricornutum (b). Error bars represent SE (n ≥ 3).

Rate constant of protein turnover

The 3H2O specific activity used in this study did not cause any detectable radiation stress or damage to cells. Growth rates in labelled and unlabelled cells were similar when a 20-fold range (1.85–37 kBq mL−1) of 3H2O was used on both species grown at µmin and µmax (Quigg 2000). Dunaliella tertiolecta and P. tricornutum grown at µmax with 18.5 kBq mL−1 3H2O maintained a relatively constant Rd : Pm ratio for the duration of the chase period, which was similar to that in control chambers and turbidostats (Table 2).

Table 2.  Average Rd : Pm ratios (± SE) for D. tertiolecta and P. tricornutum grown at µmax in turbidostats (n ≥ 12) and chambers with and without 3H2O (n ≥ 20)
 Rd : Pm ratio TurbidostatsRd : Pm ratio in chambers:
D. tertiolecta0.230 (0.035)0.259 (0.027)0.299 (0.029)
P. tricornutum0.228 (0.018)0.267 (0.025)0.271 (0.019)

3H completely labelled a small pool of rapidly turning over protein within the first 2 h of the pulse period (Fig. 1). Average incorporation rates were similar for D. tertiolecta (18 ± 2 × 10−4 mol 3H g protein−1 h−1; n = 10) and P. tricornutum (17 ± 1 × 10−4 mol 3H g protein−1 h−1; n = 11). This pool seemed to account for approximately 0.4% of the total protein in the cell (Quigg 2000). Thereafter, labelling showed a gradual increase suggesting slow entry of label into a larger protein pool (Fig. 1). This pool, however, did not appear to become fully labelled during the 24 h pulse, but presumably (ignoring the possibility of multiple pools with different turnover times) was labelled sufficiently such that a long-term chase would still yield good estimates of the kinetics of long-term protein turnover.

Short-term protein turnover rates were measured in the first 10 h of the chase (Fig. 5a) whereas long-term protein turnover rates were measured over a 100 h chase (Fig. 5b). Rate constants for protein degradation at each photon flux were faster in D. tertiolecta than in P. tricornutum (Fig. 5a & b). Protein turnover rates increased with growth rate. The long-term protein turnover rates were found to be significantly different (P < 0.05) when µmin and µmax were compared. Short-term protein turnover rates however, did not vary significantly between treatments (P > 0.05).

Figure 5. Short‐term (a) and long‐term (b) rate constants for protein turnover in D. tertiolecta (dashed line) and P. tricornutum (solid line) were plotted against the photon flux.

Figure 5. Short-term (a) and long-term (b) rate constants for protein turnover in D. tertiolecta (dashed line) and P. tricornutum (solid line) were plotted against the photon flux.

for growth. Mean rates are presented with SE (n > 3). Linear regressions describing this relationship are: D. tertiolecta: y = 0.007x + 2.12, r2 = 0.99 (a), y = 0.001x + 0.17, r2 = 0.78 (b); P. tricornutum: y = 0.011x + 3.18, r2 = 0.79 (a), y = 0.001x + 0.32, r2 = 0.66 (b).

Maintenance coefficient for protein turnover

The maintenance coefficient for protein turnover (mp′) is the respiratory cost on a protein-N and time basis (mol O2 mol−1 N s−1). The values of mp′ for D. tertiolecta and P. tricornutum were estimated according to Eqn 1 (De Visser et al. 1992; Van der Werf et al. 1992; Bouma et al. 1994).

mp′  =  Esp  ×  kd  ×  k(1)

Rate constants for protein degradation (kd, d−1) were taken from the zero growth rate intercept on Fig. 5b. The minimum theoretical specific ATP cost for protein turnover (Esp) is 10.8 mol ATP mol peptide bond−1 (Raven et al. 2000). A correction factor (k) was used to transform units of Esp and kd (De Visser et al. 1992). For direct comparisons of mp′ with maintenance metabolic rates (below) and in the literature, we also then assumed that protein-N accounted for 85% of cellular-N (De Visser et al. 1992) and used the protein pool sizes at Icg (Table 1). On this basis, values for mp′ were 6.37 × 10−6 mol O2 mol−1 C s−1 (0.55 d−1) and 0.84 × 10−6 mol O2 mol−1 C s−1 (0.07 d−1) for D. tertiolecta and P. tricornutum, respectively.



Both laboratory experiments and field observations suggest that microalgae are capable of growth at extremely low photon fluxes, ≤ 1 µmol photon m−2 s−1 (Richardson, Beardall & Raven 1983; Geider et al. 1985, 1986; Falkowski & Raven 1997; Raven et al. 2000). Strategies employed are species-specific as seen in Figs 2–4 and Tables 1 and 2. Phaeodactylum tricornutum was able to sustain growth at photon fluxes an order of magnitude lower than D. tertiolecta (Fig. 2a & b), which involved lowered rates of RdC relative to those measured in D. tertiolecta (Fig. 3a & b; Table 1). Geider et al. (1985) have previously reported that RdC did not change significantly as a function of the photon flux for growth in P. tricornutum, with similarly elevated RdC at low photon fluxes for growth (Fig. 3b). We suggest that higher than predicted RdC may provide this alga with respiratory-derived ATP and reductant for continued growth in the virtual absence of photosynthesis, and hence be an important factor allowing this alga to successfully exploit variable environments.

An increase in RdC and PmC with photon flux for growth is considered the typical response for microalgae (Richardson et al. 1983; Langdon 1988; Geider & Osborne 1989) and higher plants (Lambers et al. 1998). This was observed for D. tertiolecta up to approximately 200 µmol photons m−2  s−1, after which there was no real change in RdC (Fig. 3a) but a small decline in PmC (Table 1), suggesting that D. tertiolecta may have been photo-inhibited at these higher photon fluxes. Close to Icg, C and N quotas rose significantly in D. tertiolecta (Fig. 4a). Cells were seen to accumulate starch (Quigg 2000) and increase the size of their protein pool (Table 1). Despite these physiological strategies, D. tertiolecta was not able to grow at photon fluxes as low as those exploited by P. tricornutum.

Protein turnover in eukaryotic microalgae

We still do not have a clear understanding of the in vivo rates, mechanisms and energy costs of biosynthesis and biodegradation of proteins and regulation of enzyme activity in microalgae. Long-term rate constants of protein turnover in D. tertiolecta and P. tricornutum ranged from 0.14 to 0.29 d−1 and 0.29–0.45 d−1, respectively. These are comparable with rates of protein turnover in Chlorella fusca var. vacuolata and higher plants (Table 3). Protein turnover rates in leaves and roots are species-specific, and depend on both the growth rate and environmental conditions. Richards & Thurston (1980a, b) found that protein turnover rates in C. vacuolata decreased by 30% upon nitrogen starvation. Similar decreases in the rate of protein turnover have been measured in higher plants (Table 3). Therefore, as the growth rate decreases, so does the overall rate of protein turnover in bacteria, microalgae and higher plants.

Table 3.  Reported measurements of the protein turnover rate (d−1) and of the respiratory energy demand (%) associated with protein turnover in higher plants
SpeciesCharacteristicsRate of protein turnover (d−1)% of maintenance respiration ascribed to protein turnoverReference
  1. aPenning de Vries (1975), Davies (1982) are reviews.

  2. bvan der Werf et al. (1992) and Scheurwater et al. (2000) assumed 50% recycling of the 14C-labelled leucine in their experiments.

  3. nd, not done

Higher plantsLeaves0.1250–60%Penning de Vries (1975)a
Lemna minorLeaves0.09 non-stressedndHumphrey & Davies (1976)
0.34 stressed  
Chlorella fusca var. vacuolataChlorophyte0.30–0.51 non-stressedndRichards & Thurston (1980a, b)
Zea maysLeaves0.12ndSimpson, Cooke & Davies (1981)
Hordeum vulgareLeaves, seedlings0.09 non-stressedndDungey & Davies (1982)
0.15 stressed  
Higher plantsLeaves0.42ndDavies (1982)a
Lolium perenneLeaves0.2827–36%Barneix et al. (1988)
Dactylis glomerataRoots0.0657%van der Werf et al. (1992)b
Phaseolus vulgaris
Solanum tuberosum
Lolium perenne
Hordeum vulgareLeaves, full-grownnd30–60%De Visser et al. (1992)
Solanum tuberosumLeaves, growing0.04–0.2117–35%Bouma et al. (1994)
Phaseolus vulgarisLeaves, full-grown0.04–0.09  
Triticum aestivumLeavesnd34–37%Zagdañska (1995)
Dactylis glomerateRoots, fast growing0.15622–30%Scheurwater et al. (2000)b
Festuca ovinaRoots, slow growing0.11622–30% 

Our estimates of protein turnover assumed synthesis and degradation occurred simultaneously. As such, we did not expect to see a net change in the cellular protein concentrations. In line with this, the protein pool size essentially did not vary in response to the photon flux for growth in D. tertiolecta and P. tricornutum (with only one exception; Table 1). Protein pool sizes have been shown to remain relatively constant as a function of the growth irradiance in other eukaryotic microalgae: Thalassiosira weisflogii (Post, deWit & Mur 1985), Chaetoceros gracilis, Isochrysis aff. galbana, P. tricornutum (Thompson, Guo & Harrison 1993) and Leptocylindrus danicus (Verity 1982). Although the size of the protein pool per se remained unchanged, the amounts of specific proteins and the types of proteins synthesized at the extreme ranges of photon fluxes for growth would have varied. For example, light-harvesting complex proteins are synthesized to a greater extent in low-light-grown cells, while the opposite occurs for Rubisco (Falkowski & Raven 1997).

Proportion of protein turnover related to maintenance costs

Protein turnover is considered to be the largest single contribution to maintenance costs in all living cells (Penning de Vries 1975; Raven & Beardall 1981; Geider et al. 1996; Lambers et al. 1998; Raven et al. 2000). In order to determine what proportion of respiration was associated with protein turnover, the specific maintenance respiration rate (Rm) is calculated. Langdon (1988) and Geider & Osborne (1989) reviewed estimates of Rm calculated when light was the limiting factor for growth in 17 species of microalgae. These varied between 0.11 and 5.09 × 10−6 mol O2  mol−1 C s−1 (i.e. 0.009–0.44 d−1 assuming a respiratory quotient of unity). However, the question of which is the most appropriate method to calculate Rm remains contentious.

Extrapolation of the growth rate versus irradiance relationship (Fig. 2a & b) to zero growth as proposed by Van Liere & Mur 1979) yielded a specific maintenance coefficient of 0.33 d−1 for D. tertiolecta and 0.27 d−1 for P. tricornutum. The intercept of the linear portion of the relationship between RdC and growth rate has also been used to estimate Rm (Laws et al. 1985). We did this only for D. tertiolecta (Fig. 3a), and calculated a value for Rm of 0.056 mol O2 mol−1 C d−1. The model used by Geider et al. (1985, 1986) and Langdon (1988) is the method we preferred to estimate Rm, as this considers a number of physiological attributes that constrain growth (Eqn 2),

Rm  =  [a* φm Icg Θ Z]/PQ(2)

and hence the magnitude of Rm. The Chl a-specific absorption coefficients (a*) were taken from the literature: 0.0052 m2 mg−1 Chl a for D. tertiolecta (Welschmeyer & Lorenzen 1981 in Langdon 1988) and 0.0051 m2 mg−1 Chl a for P. tricornutum (Geider et al. 1985). The maximum theoretical quantum yield of photosynthesis (φm) is 0.125 mol O2 mol−1 photon. The Chl a : C ratios (Θ; Table 1) at Icg were 0.086 and 0.056 mg Chl a mg−1 C for D. tertiolecta and P. tricornutum, respectively. Z is the atomic weight of carbon (0.012 mg mol C−1). Photosynthetic quotients (PQ) for algae have been shown to depend on the nitrogen source during growth (Laws 1991). As we used nitrate in the growth media, we used a value for PQ of l.4 mol O2 mol−1 CO2. The Rm calculated for D. tertiolecta (8.62 × 10−6 s−1; 0.87 d−1) was greater than that of P. tricornutum (0.73 × 10−6 s−1; 0.06 d−1). The Rm for P. tricornutum was similar to that previously reported by Geider et al. (1985) for the same species. Both our mp′ (Eqn 1) and Rm values (Eqn 2) were an order of magnitude greater for D. tertiolecta than for P. tricornutum. These results are consistent with the literature which proposes that species that are best adapted to growth at low light exhibit lower maintenance respiration rates (Langdon 1988), with diatoms typically exhibiting lower Rm values than flagellates (Falkowski & Raven 1997). Variations in the magnitude of Rm calculated using the various models most likely reflects the associated assumptions.

Using the Rm calculated from Eqn 2, we estimated the proportion of maintenance respiration used for protein turnover (Table 4). Protein turnover was proportionately a more energetically expensive process in D. tertiolecta (36%) than in P. tricornutum (30%). These values fall within the reported range of 7–60% for higher plants (Table 3). The trade-off for high metabolic activity (requiring high protein concentrations) as seen in D. tertiolecta is an associated increase in the cost of maintenance. Higher respiration rates as measured in D. tertiolecta contributed to higher Icg values, which, in turn, increased the minimum irradiance at which this alga was able to maintain a positive carbon balance. Hence, the contention that any factor that increases protein concentration or turnover rate will increase the maintenance respiration in higher plants (Penning de Vries 1975; Davies 1982; Vierstra 1993) is supported by the findings for these two species of eukaryotic microalgae.

Table 4.  An estimate of the proportion of maintenance respiration used for protein turnover by D. tertiolecta and P. tricornutum. The rate constant for protein turnover was measured at the zero growth rate intercept in Fig. 5a and b. The magnitude of the maintenance metabolic rate was estimated according to Eqn 2 (Geider et al. 1985, 1986; Langdon 1988). An ATP demand of 10.8/peptide bond is based on Raven et al. (2000) and the mean molecular weight per amino acid (110 Da) is from De Visser et al. (1992). The equivalent O2 consumption rate was calculated by Penning de Vries (1975)
Estimate of the maintenance respiration rateD. tertiolectaP. tricornutum
mol O2 mol C−1 s−1 (From Eqn 2)8.62 × 10−60.73 × 10−6
Carbon content (mol C cell−1) at Icg4.83 × 1012 1.0 × 1012
mol O2 cell−1 h−16.40 × 10−152.63 × 10−15
Respiration rates based on protein turnoverLong-termShort-termLong-termShort-term
  • a

    Short-term protein turnover costs as a percentage of maintenance respiration are adjusted, based on the estimation that only 0.4% of the protein pool is turned over rapidly.

Rate constant for protein turnover at zero growth rate (h−1)0.0070.090.0130.13
Rate of turnover (g protein cell−1 h−1)1.16 × 10−1314.4 × 10−130.40 × 10−134.03 × 10−13
ATP demand at 10.8/peptide bond and mean molecular weight/amino acid of 110 Da (mol ATP cell−1 h−1)1.14 × 10−1413.96 × 10−130.39 × 10−143.95 × 10−14
Equivalent O2 consumption rate assuming 0.5 O2 → H2O with translocation of 10 H+/ATP (mol O2 h−1 cell−1)2.29 × 10−152.79 × 10−140.79 × 10−150.79 × 10−14
Protein turnover as a percentage of maintenance respiration36%1.7%a30%1.2%a

As nothing is known about the relative contribution of the protein pools with high and low turnover rates, calculations presented here have been based on the lowest protein turnover rates measured. The phenomenon for the short-term protein turnover rates is not so easily explained by our current understanding of the processes that govern protein turnover and maintenance. However, by using our estimate for the rapidly turned-over pool size (0.4%), we predict that short-term protein turnover rates account for only approximately 1.7 and 1.2% of the total maintenance cost in D. tertiolecta and P. tricornutum, respectively. Hence, short-term protein turnover rates will not influence cellular energy budgets, even at very low photon fluxes.

Species-specific ranges of protein turnover rates add to the known list of differences in the photo-acclimation strategies used by D. tertiolecta and P. tricornutum. The overall higher protein turnover rates and two-fold range measured in D. tertiolecta would be important in dealing with the potentially damaging effects of higher photon fluxes, as well as increasing protein pool sizes when growing at extremely low photon fluxes. The lowered maintenance metabolic energy requirement of P. tricornutum for cellular housekeeping, is arguably the most important factor governing its ability to grow at very low photon fluxes, along with the higher than predicted RdC. The present investigation examined the impact of protein turnover as a constraint limiting to photolithotrophic growth at low light. The two other constraints proposed by Raven et al. (2000) – charge recombination in photosystem II and proton leakage and slippage – are the subject of future investigations.


A.Q. would sincerely like to thank John Raven for his continued support. Karen Kevekordes, Zoe Finkel and Oscar Schofield read and provided useful comments on earlier drafts of this manuscript. Andrew Johnston of the Department of Biological Sciences, Dundee University, Scotland, generously performed the CHN analyses. We would like to thank two anonymous reviewers for helpful comments.

Received 21 July 2002; received in revised form 25 October 2002; accepted for publication 29 October 2002