Growth, lipid accumulation, and fatty acid composition in obligate psychrophilic, facultative psychrophilic, and mesophilic yeasts


  • Editor: Max Häggblom

Correspondence: Maddalena Rossi, Department of Chemistry, University of Modena and Reggio Emilia, via Campi 183, 41100 Modena, Italy. Tel.: +39 059 205 5567; fax: +39 059 373 543; e-mail:


Obligate psychrophilic, facultative psychrophilic, and mesophilic yeasts were cultured in a carbon-rich medium at different temperatures to investigate whether growth parameters, lipid accumulation, and fatty acid (FA) composition were adaptive and/or acclimatory responses. Acclimation of facultative psychrophiles and mesophiles to a lower temperature decreased their specific growth rate, but did not affect their biomass yield (YX/S). Obligate and facultative psychrophiles exhibited the highest YX/S. Acclimation to lower temperature decreased the lipid yield (YL/X) in mesophilic yeasts, but did not affect YL/X in facultative psychrophilic ones. Similar YL/X were found in both groups of psychrophiles, suggesting that lipid accumulation is not a distinctive characteristic of adaptation to permanently cold environments. The unsaturation of FAs was one major adaptive feature of the yeasts colonizing permanently cold ecosystems. Remarkable amounts of α-linolenic acid were found in obligate psychrophiles at the expense of linoleic acid, whereas it was scarce or absent in all the other strains. Increased unsaturation of FAs was also a general acclimatory response of facultative psychrophiles to a lower temperature. These results improve the knowledge of the responses enabling psychrophilic yeasts to cope with the cold and may be of support for potential biotechnological exploitation of these strains.


Permanently cold ecosystems make up one of the largest biospheres on the Earth and represent one of the last unexplored frontiers of ecology. Psychrophilic microorganisms colonize deep oceans, glaciers, and polar areas and constitute an important part of cold-adapted biodiversity, playing an essential role as nutrient cyclers and waste mineralizers. Conventionally, obligate psychrophiles have a maximum temperature for growth <20 °C, an optimum temperature of c. 15 °C, and a minimum temperature at 0 °C or lower. On the contrary, facultative psychrophiles can be thought of as mesophiles that evolved to tolerate cold. They exhibit optimum temperatures >20 °C and are capable of growth around 0 °C (Cavicchioli & Tortsen, 2000; Price & Sowers, 2004; Raspor & Zupan, 2006; Margesin et al., 2007b). Even though psychrophilic microorganisms are still understudied, they are increasingly being investigated for valuable biotechnological purposes (Margesin & Schinner, 1999a, b; Russell, 2000; Margesin et al., 2007b, c).

To successfully colonize cold environments, psychrophiles have evolved a number of genetic-based phenomena, ascribed to adaptive and acclimatory responses. The time-scale exposure is essential to interpret the effects of suboptimal temperature on microbial physiology. An abrupt temperature shift is likely to generate rapid, high-dynamic stress-response phenomena on microbial cells, whereas prolonged exposure to suboptimal temperature leads to acclimation, which implicates regulatory mechanisms resulting in the full adjustment of the genomic expression and the physiological state during a lifetime. Over a time scale of several generations, the evolutionary selection of the gene alleles, which increase the fitness in a specific environmental niche, is termed adaptation (Morgan-Kiss et al., 2006).

Yeasts inhabit diverse ecological niches including those that are considered extreme with respect to low temperatures (Larkin & Stokes, 1968). A few studies reported the isolation of yeasts in meltwater rivers originating from the glaciers of Patagonia (de García et al., 2007), in the different ice layers of Arctic glaciers (Gostinčar et al., 2006), in sediments, ice, meltwaters, and soils of alpine environments (Margesin et al., 2007a; Turchetti et al., 2008). Although the ability of yeasts to cope with low environmental temperatures has attracted considerable attention, so far, the mechanisms underlying this phenomenon have not been described fully for psychrophilic yeasts, because most of the studies have been focused on mesophilic ones, particularly on Saccharomyces cerevisiae (Sahara et al., 2002; Homma et al., 2003; Schade et al., 2004; Aguilera et al., 2007; Tai et al., 2007). A temperature decrease from 30 to 12 °C affects the transcription level of many genes of S. cerevisiae, such as the ones involved in the regulation of transporters of growth-limiting nutrients, glycogen metabolism, ribosome biogenesis, and ribosomal proteins. Trehalose biosynthesis is involved in cold shock response, but not in steady-state low-temperature acclimation. A defined group of genes regulating the response of this yeast species to cold is involved in lipid metabolism. Δ9 Desaturase, which is the key enzyme in the synthesis of unsaturated fatty acyl-CoAs and is the only known desaturase of S. cerevisiae, is overexpressed at low temperatures (Stukey et al., 1989; Nakagawa et al., 2002; Martin et al., 2007).

Changes in lipid metabolism are known to constitute the major adaptation of metabolic functions occurring during growth at low temperatures. They determine an improved membrane fluidity causing both the maintenance of the appropriate physical state of lipid bilayer and the good functionality of membranes (Los & Murata, 2004; Morgan-Kiss et al., 2006). Differences in fatty acid (FA) composition have been described since the 1970s in yeasts growing at diverse temperatures (McMurrough & Rose, 1973; Arthur & Watson, 1976; Watson & Arthur, 1976; Watson, 1978; Watson et al., 1978), but no rigorous comparative studies of the physiological behavior of obligate psychrophilic, facultative psychrophilic, and mesophilic yeast strains have been published yet. The present study wanted to fill this gap by comparing the growth parameters and FA composition of 26 yeasts strains (representative of 12 species) isolated from different cold and temperate habitats and clustered as obligate psychrophiles, facultative psychrophiles, and mesophiles.

Materials and methods


All chemicals were obtained from Sigma-Aldrich (Steinheim, Germany) unless otherwise stated.

Yeast strains

The yeast strains used in this study were obtained from the Industrial Yeasts Collection DBVPG (University of Perugia, Italy,, from the American Type Culture Collection (ATCC), or from our own collection. Twenty psychrophilic yeasts were isolated from sediments, ice, and meltwater of alpine glaciers (Turchetti et al., 2008). Twelve strains belonged to the species Cryptococcus gilvescens, Rhodotorula creatinivora, Rhodotorula laryngis, Rhodotorula glacialis, and the yeast-like species Aureobasidium pullulans. Besides, eight strains were preliminarily identified as Leucosporidium spp. and Mrakia spp., because they belong to two so far undescribed new species. Accordingly, their formal taxonomic description is still in progress.

The six mesophilic strains, which were isolated from temperate habitats and exhibited temperature optima ranging between 25 and 35 °C, belonged to the species S. cerevisiae, Saccharomyces exiguus, Kluyveromyces marxianus, Pichia farinosa, and Zygosaccharomyces rouxii.

Growth experiments

All the strains were aerobically cultured in flasks containing the carbon-rich medium GMY (Buzzini, 2001): 40 g L−1 glucose, 8 g L−1 KH2PO4, 0.5 g L−1 MgSO4·7H2O, and 3 g L−1 yeast extract (Difco Laboratories, Sparks, MD), pH 5.5. The yeasts from cold habitats were subcultured at 4 °C for 14 days, while those from temperate habitats were subcultured at 30 °C for 5 days.

To determine the growth capabilities at different temperatures, batch cultures were carried out in shake flasks containing 50 mL GMY broth. The flasks were inoculated (10% v/v) with grown seed cultures and were incubated at 4, 18, or 30 °C. The turbidity (OD600 nm) was monitored during batch cultivation to calculate the specific growth rate (μ) and to determine the growth kinetic. The residual glucose concentration, the biomass dry weight, the viable count on GMY agar plates, and lipid composition were determined at the stationary phase (after 14 days for cultures grown at 4 and 18 °C, and after 5 days for cultures grown at 30 °C).

Biomass dry weight was determined gravimetrically. Glucose was analyzed by means of HPLC equipped with an RI detector (HPLC System, 1200 Series, Agilent Technologies, Santa Clara, CA). The analysis was performed using an Aminex HPX-87H ion exclusion column (300 × 7.8 mm) and 5 mM H2SO4 (flow rate 0.6 mL min−1) as the isocratic mobile phase. The biomass/substrate yield (YX/S) was expressed as g dry biomass g−1 of consumed glucose.

Lipid analysis

Biomass from 50-mL culture samples was harvested (3000 g, 5 min, 0 °C), washed with distilled water, frozen at −80 °C, and lyophilized (Lyolab 3000, Heto-Holten, Allerød, Denmark). Lipids were extracted using the procedure of Folch et al. (1957), with a few modifications. One gram of lyophilized biomass was extracted with 50 mL of a chloroform : methanol mixture (2 : 1, v/v) under shaken conditions for 16 h. The extract was filtered through a column of celite and anhydrous Na2SO4 to remove cell debris and water. Solvents were removed with a rotavapor and then lipids were weighed. The lipid/biomass yield (YL/X) was expressed as g lipid extract g−1of dry biomass.

The extract was subjected to methanolysis of total lipids according to the method of Morrison & Smith (1964). Lipids were dissolved in 2 mL of a 1 : 1 mixture of hexane and BF3 solution (14% in methanol, Sigma-Aldrich) and transferred into a Schlenk tube. Five milligrams of glyceryl triundecanoate was added to the reaction mixture to generate the internal standard for GC-MS analysis. Transesterification was carried out at 100 °C for 1 h, and then 2 mL of d.d. water was added to quench the reaction.

The organic phase was collected and the fatty acyl methyl esters were analyzed by GC. Analysis was performed using a quadrupole GC-MS system (HP5890 Series II gas chromatograph – HP5972 mass selective detector) equipped with an EI ionization detector (70 eV ionization energy). An HP-5 capillary column was used for the separation (Agilent Technologies, internal diameter of 0.20 mm, film thickness of 0.5 μm, and length of 30 m). The injection temperature was 280 °C and the oven temperature was programmed from 80 °C (1-min isotherm) to 130 °C at a rate of 50 °C min−1, and then to 280 °C at a rate of 5 °C min−1 (20-min isotherm at 280 °C). High-purity helium was used as the mobile phase and a constant column head pressure of 9 psi was maintained during the analyses.

FAs were identified by comparison with commercial standards of fatty acyl methyl esters. Peak areas in the total ion chromatograms were used to determine their relative amounts. The quantitative equation of the unsaturation index (UI) was used to aid the analysis of lipid composition. The UI was calculated as the number of double bonds of each FA multiplied by its relative amount (Watson & Arthur, 1976).

Statistical analysis

All values are means of three separate experiments. Differences in means among groups A, B, and C were evaluated using one-way anova, followed by Tukey's post hoc comparisons. Differences were considered statistically significant at P≤0.05. Differences in means among the growth temperatures were analyzed using two-way anova with repeated measures with the group as the first factor and temperature as the second factor, followed by Bonferroni's post hoc comparisons. Differences were considered statistically significant at P≤0.05. Statistical analysis was performed using graphpad prism 4.0 (Graphpad Software, San Diego, CA).


Growth parameters of facultative and obligate psychrophilic, and mesophilic yeasts

The ability of all the yeasts to grow at 4, 18, and 30 °C was determined (Table 1). All the strains from cold habitats grew at 4 °C, but they were divided into obligate and facultative psychrophiles (groups A and B, respectively) on the basis of their ability to grow at 18 and 30 °C. The strains belonging to the species R. creatinivora, R. glacialis, Leucosporidium sp., and Mrakia sp. did not grow at 18 and 30 °C and behaved as obligate psychrophiles (group A). The strains belonging to the species R. laryngis, C. gilvescens, and A. pullulans grew at 18 and 30 °C and were grouped as facultative psychrophiles (group B) (Table 1). All the strains from temperate habitats (group C) were successfully cultured at both 18 and 30 °C, but only S. cerevisiae L12, S. cerevisiae ATCC 2345, and K. marxianus L3 grew at 4 °C (Table 1).

Table 1.   Specific growth rate (μ), biomass/substrate yield (YX/S), and lipid/biomass yield (YL/X) of obligate psychrophilic (group A), facultative psychrophilic (group B), and mesophilic (group C) yeasts at 4, 18, and 30°C
Strainμ (h−1)*YX/SYL/X
  • *

    Mean values from three experiments; within each strain and condition SD always <0.005 h−1.

  • Mean values from three experiments; within each strain and condition SD always <0.05.

  • Means in a column without a common letter differ (P<0.05). Means in a row without a common symbol differ (P<0.05).

Group A
Leucosporidium sp. DBVPG 47530.0220.300.28
Leucosporidium sp. DBVPG 47990.0320.340.58
Rhodotorula creatinivora DBVPG 47940.0350.330.46
Mrakia sp. DBVPG 47750.0390.370.25
Mrakia sp. DBVPG 47560.0350.320.17
Mrakia sp. DBVPG 47570.0330.330.34
Mrakia sp. DBVPG 47610.0540.340.34
Mrakia sp. DBVPG 47540.0360.320.19
Mrakia sp. DBVPG 48010.0380.330.13
Rhodotorula glacialis DBVPG 48060.0510.290.42
Rhodotorula glacialis DBVPG 47850.0530.450.21
Rhodotorula glacialis DBVPG 47860.0460.380.17
Rhodotorula glacialis DBVPG 47160.0360.440.08
Mean0.039b  0.35b  0.28b  
Group B
Rhodotorula laryngis DBVPG 47650.0370.0600.1020.490.440.450.070.090.05
Rhodotorula laryngis DBVPG 47720.0250.0350.1280.470.430.490.280.340.04
Aureobasidium pullulans DBVPG 47780.0450.0640.1090.270.280.330.280.120.08
Cryptococcus gilvescens DBVPG 47140.0320.1110.1060.
Cryptococcus gilvescens DBVPG 47200.0370.1160.1180.290.290.300.210.270.30
Cryptococcus gilvescens DBVPG 48030.0460.0870.0870.360.
Cryptococcus gilvescens DBVPG 47220.0530.0770.0820.350.270.340.260.360.26
Group C
Saccharomyces cerevisiae L 120.0260.1400.2490.
Saccharomyces cerevisiae ATCC 23450.0270.1510.2670.
Saccharomyces exiguus L 100.1470.3230.070.070.720.51
Klyveromyces marxianus L 30.0310.1750.2330.
Pichia farinosa DBVPG 36260.1710.3050.190.180.570.29
Zigosaccharomyces rouxii DBVPG 63990.0990.1920.180.110.430.31

The specific growth rate (μ) and the biomass/substrate yield (YX/S) of all the yeasts were determined at the different growth temperatures (Table 1). Both obligate and facultative psychrophilic yeasts (groups A and B) grew at 4 °C with a mean μ of 0.039 h−1. This value was significantly higher than that exhibited by the three mesophiles that grew at this temperature (0.028 h−1) (Table 1). When the yeasts of groups B and C were cultured at 18 and 30 °C, their mean μ significantly increased as the temperature was increased. At the same time, the strains belonging to group B grown both at 18 and 30 °C showed a mean μ significantly lower than that exhibited by group C under the same conditions.

The conversion of glucose into biomass (YX/S) was always significantly higher in psychrophilic (groups A and B) than in mesophilic (group C) yeasts. In both groups B and C, the growth temperature did not affect the YX/S (Table 1).

Lipid yield and composition

The lipid/biomass yield (YL/X) was determined at the stationary phase at the different temperatures (Table 1). At 4 °C, both psychrophilic yeast groups (A and B) exhibited a mean YL/X significantly higher than that showed by the three mesophilic strains growing at the same temperature. Yeasts of group B displayed no significant differences in the mean YL/X at 4, 18, and 30 °C. In contrast, the YL/X increased in strains belonging to group C from 4 to 18 °C and 30 °C (Table 1).

The relative composition of FAs was determined in the lipid fraction of biomass at the stationary phase (Table 2). Only linear FAs were found in all the yeasts, independent of the growth temperatures. Saturated and unsaturated FA with chain length ranging from 14 to 18 carbons always accounted for >97%. C14 FA never exceeded 5.3%. The relative amount of unsaturated FA other than palmitoleic (C16:1, Δ9), oleic (C18:1, Δ9), linoleic (C18:2, Δ9,12), and α-linolenic (C18:3, Δ9,12,15) was negligible.

Table 2.   Relative composition (%) of fatty acids in the lipid extracts of obligate psychrophilic (group A), facultative psychrophilic (group B), and mesophilic (group C) yeasts at stationary phase at 4, 18, and 30°C
  • *

    Mean values from three experiments; within each strain and condition SD always <0.5.

  • Means in a column without a common letter differ (P<0.05). Means in a row without a common symbol differ (P<0.05).

Group A
Leucosporidium sp. DBVPG 47531.618.53.00.358.312.94.9
Leucosporidium sp. DBVPG 47991.
R. creatinivora DBVPG 47942.625.53.01.952.53.411.0
Mrakia sp. DBVPG 47750.315.73.54.728.717.829.1
Mrakia sp. DBVPG 47560.619.
Mrakia sp. DBVPG 47570.
Mrakia sp. DBVPG 47610.
Mrakia sp. DBVPG 47540.718.33.26.827.416.726.6
Mrakia sp. DBVPG 48010.
R. glacialis DBVPG 48065.322.12.51.445.411.511.1
R. glacialis DBVPG 47854.814.51.42.639.421.215.4
R. glacialis DBVPG 47862.
R. glacialis DBVPG 47160.325.51.810.843.414.24.1
Mean1.5ab  19.3a  2.3b  3.8a  41.1b  14.9b  16.8b  
Group B
R. laryngis DBVPG 47650.80.01.626.023.320.
R. laryngis DBVPG 47720.
A. pullulans DBVPG 47780.00.00.922.514.
C. gilvescens DBVPG 47140.
C. gilvescens DBVPG 47202.
C. gilvescens DBVPG 48030.
C. gilvescens DBVPG 47220.00.00.317.120.714.
Group C
S. cerevisiae L
S. cerevisiae ATCC 23453.
S. exiguus L 103.64.328.920.321.935.415.714.
K. marxianus L
P. farinosa DBVPG 36260.70.920.423.
Z. rouxii DBVPG 63991.71.216.917.512.19.911.211.538.852.318.

Significant (P>0.05) differences were observed in the unsaturated FA composition among groups A, B, and C at 4 °C (Table 2). At this temperature, the relative amount of C18:1 and C18:2 in psychrophilic strains (groups A and B) was significantly higher than that observed in the yeasts of group C. Besides, the relative composition of C18:2 significantly differed between groups A and B (14.9% and 24.5%, respectively). Moreover, obligate psychrophilic strains (group A) exhibited a significantly (P>0.05) higher concentration of C18:3 (16.8%) than those observed in the other two groups (B and C). Among the yeasts belonging to group C, only K. marxianus L3 produced C18:2 or C18:3 at 4 °C (Table 2). On the other hand, at 4 °C, strains belonging to group C showed a significantly (P>0.05) higher relative percentage of C16:1 (39.7%) than those observed in yeasts of groups A and B (2.3% and 0.4%, respectively). No clear tendency was observed in terms of the relative concentration of saturated FA (C14, C16, and C18) among the three groups (Table 2).

Strain-to-strain differences in the FA composition did not allow distinguishing any common trend as a function of growth temperature within groups B and C. In group B, the sole statistically significant difference was the content of C18:1, which was higher at 18 and 30 °C (58.4% and 57.6%, respectively) than at 4 °C (41.7%). Likewise, a similar behavior was observed for strains of group C, which exhibited an increase of both C18 and C18:1 from 4 °C to 18 and 30 °C. These same strains exhibited an increase of C16 and a simultaneous decrease of C16:1 when the growth temperature increased from 4 to 18 °C and 30 °C (Table 2).

In order to aid comparison, information about unsaturation was summarized as the UI (Table 3). The different growth temperatures at 4, 18, and 30 °C did not significantly affect the UI of strains belonging to groups B and C. Obligate psychrophilic yeasts (groups A) exhibited a significantly (P>0.05) higher UI at 4 °C than those of groups B and C. By comparing the data reported in Table 3 with those reported in Table 2, it could be observed that the different UI between groups A and B was mainly due to the higher amount of C18:3. On the other hand, the lower UI of yeasts belonging to group C was mainly due to the general lower content of unsaturated C18:1, C18:2, and C18:3, which was only partially balanced by a higher content of C16:1. On the whole, the major contribution to the UI provided by strains of group C was due to C16:1 and C18:1, whereas in psychrophilic yeasts (groups A and B) the main contribution was due to C18:1, C18:2, and C18:3 (Table 2).

Table 3.   UI and relative composition (%) of saturated, MUFAs, PUFAs, total C16, and total C18 fatty acids in the lipid extracts of obligate psychrophilic (group A), facultative psychrophilic (group B), and mesophilic (group C) yeasts at 4, 18, and 30°C
StrainUI*SaturatedMUFAsPUFAsTotal C16Total C18
  • *

    Mean values from three experiments; within each strain and condition SD always <0.04.

  • Mean values from three experiments; within each strain and condition SD always <2.

  • Means in a column without a common letter differ (P<0.05). Means in a row without a common symbol differ (P<0.05).

Group A
Leucosporidium sp. DBVPG 47531.0220.461.317.821.576.4
Leucosporidium sp. DBVPG 47991.0618.762.718.619.379.7
R. creatinivora DBVPG 47940.953055.514.428.568.8
Mrakia sp. DBVPG 47751.5520.732.246.919.280.3
Mrakia sp. DBVPG 47561.4628.625.944.921.277.6
Mrakia sp. DBVPG 47571.2123.846.629.621.778.3
Mrakia sp. DBVPG 47611.5018.139.442.519.880
Mrakia sp. DBVPG 47541.4425.830.643.321.577.5
Mrakia sp. DBVPG 48010.9225.260.614.326.373.8
R. glacialis DBVPG 48061.0528.847.922.624.669.4
R. glacialis DBVPG 47851.3021.940.836.615.978.6
R. glacialis DBVPG 47861.7821.516.261.914.682.8
R. glacialis DBVPG 47160.8636.645.218.327.372.5
Mean1.24b  24.6a  43.5a  31.7a  21.6a  76.6b  
Group B
R. laryngis DBVPG 47651.031.060.9528.025.426.541.245.852.230.828.821.327.823.321.471.476.777.0
R. laryngis DBVPG 47720.970.900.9531.431.431.842.346.741.426.121.826.823.331.026.375.668.073.7
A. pullulans DBVPG 47780.911.060.8426.615.930.957.962.054.015.422.215.122.515.128.577.485.070.6
C. gilvescens DBVPG 47140.900.710.9442.331.421.224.164.863.632.
C. gilvescens DBVPG 47200.820.820.8647.530.822.922.755.968.729.512.98.526.218.518.571.380.081.5
C. gilvescens DBVPG 48031.010.741.0030.630.119.537.765.
C. gilvescens DBVPG 47220.910.830.9419.922.519.669.
Group C
S. cerevisiae L 120.740.590.7325.939.126.574.156.372.
S. cerevisiae ATCC 23450.770.620.6022.536.839.177.560.859.
S. exiguus L 100.550.6348.238.745.
K. marxianus L 31.060.980.7529.134.545.545.340.838.825.723.614.543.536.435.956.161.461.1
P. farinosa DBVPG 36260.860.7729.432.954.856.415.610.423.
Z. rouxii DBVPG 63990.890.7629.830.250.962.218.76.729.027.468.770.5

The mean relative compositions of saturated, monounsaturated fatty acids (MUFAs), polyunsaturated fatty acids (PUFAs), total C16 series (C16+C16:1), and C18 series (C18+C18:1+C18:2+C18:3) for each group grown at the diverse temperatures are reported in Table 3. At 4 °C, both psychrophilic yeast groups (A and B) exhibited a significantly higher concentration of PUFAs than those observed in group C (P>0.05). In contrast, under the same conditions, the percentage of MUFAs of group C was significantly higher than those of both groups A and B. Interestingly, facultative psychrophilic strains (group B) showed a simultaneous decrease of PUFAs and an increase of MUFAs from 4 to 18 °C and 30 °C. No similar trend of MUFAs and PUFAs as a function of the growth temperature was observed for strains belonging to group C, instead (Table 3).

Yeasts belonging to group C always displayed an average length of the FA chain lower than that found in psychrophilic ones. At 4 °C, both groups A and B exhibited a significantly higher concentration of C18 than those observed in group C (P>0.05). On the other hand, under the same conditions, strains of group C showed a higher percentage of C16 than those observed in both groups A and B (Table 3). No differences in the content of C16 and C18 were observed in group B at 4, 18, and 30 °C. In contrast, the strains of group C showed a simultaneous decrease of C18 and an increase of C16 from 4 to 18 and 30 °C.


Microbial growth is the net result of a large number of enzymatic reactions, each affected by temperature because any decrease in temperature exponentially affects the rate of biochemical reactions (Price & Sowers, 2004; D'Amico et al., 2006). As expected, the acclimation of facultative psychrophilic and mesophilic yeasts to a lower temperature affected their specific growth rates. The mean μ of both facultative psychrophiles and mesophiles was positively related to the temperature, but, unexpectedly, the growth temperature did not influence the biomass/substrate yield (YX/S) of both the groups. Obligate and facultative psychrophiles exhibited a higher efficiency of nutrient conversion into biomass (under form of YX/S) than mesophiles. Significant differences in YX/S between cold-adapted yeasts and those from temperate habitats were observed, likely resulting from adaptive responses. The dramatic gains in ATP at low and subzero temperatures observed in psycrophilic microorganisms by Napolitano & Shain (2005) apparently corroborate the above evidences.

The ability to accumulate high amounts of neutral storage lipids was reported for a few mesophilic yeasts (Gill et al., 1977; Evans & Ratledge, 1983a, b; Granger et al., 1993; Ratledge, 2002). The major neutral lipids are triacyl-glycerol and steryl-esters, which are uncharged and thus unsuitable as components of membrane lipid bilayers. Therefore, they accumulate in hydrophobic lipid particles and serve rather as storage of the building blocks for membrane lipids synthesis than as an energy reserve (Wynn et al., 2001; Wagner & Daum, 2005; Czabany et al., 2007; Daum et al., 2007). The lipid/protein ratio usually increases markedly at a low temperature in many eukaryotes and prokaryotes, due to the relative increase in lipid biosynthesis (Guschina & Harwood, 2006). Our results indicate that the acclimation of facultative psychrophilic yeasts at 30, 18, or 4 °C did not cause any significant increase in the lipid yield (YL/X). Moreover, similar YL/X were found in facultative and obligate psychrophiles at 4 °C, suggesting that lipid accumulation should not be regarded as a distinctive characteristic of adaptation to permanently cold environments.

The effects of a low temperature on the stiffening of membrane bilayers are detrimental for both eukaryotic and prokaryotic organisms. The need to acclimate and/or adapt their physiology to cold conditions is generally governed by the need to maintain membrane functionality (Los & Murata, 2004; Morgan-Kiss et al., 2006). To regulate membrane fluidity and functionality, facultative and obligate psychrophilic organisms exploit diverse changes in lipid composition, consisting in incorporation of unsaturated, short-chain, branched, or cyclic FA (White et al., 2000; Chintalapati et al., 2004; Feller, 2007). The role of unsaturation of membrane lipids is one of the most thoroughly investigated mediators of cold acclimation and/or adaptation (Russell, 1997; Guschina & Harwood, 2006; Morgan-Kiss et al., 2006). Specifically, the presence of unsaturated and PUFAs is apparently correlated with the ability of growing at a low temperature. Differences in both the FA composition and the degree of unsaturation between psychrophilic and mesophilic yeasts have been explored (McMurrough & Rose, 1973; Arthur & Watson, 1976; Watson & Arthur, 1976; Watson, 1978; Watson et al., 1978), but an extensive comparison among numerous strains of mesophilic and psychrophilic species has not been accomplished so far.

Analysis of the lipid composition of yeast strains revealed that the sum of saturated and unsaturated C14 to C18 FA always accounted for >98%. Psychrophilic yeasts differed from the mesophilic ones for their response to low temperatures. In the first case, a significant increase of PUFAs was observed when the temperature decreased to 4 °C, coupled with a parallel decrease of the MUFAs. In particular, this was the result of significantly higher amounts of linoleic (C18:2, Δ9,12) and α-linolenic (C18:3, Δ9,12,15) acids. This behavior was different from that observed in mesophilic strains, which exhibited both a significant increase of MUFAs and a decrease of PUFAs at 4 °C. In this case, this was the result of an increased concentration of palmitoleic acid (C16:1, Δ9). This dichotomy may suggest that different desaturases (specific for different yeast strains and acting on FA characterized by different length chains) could be one of the acclimation mechanisms acting to restore membrane bilayer lipid fluidity at low temperatures in either psychrophilic or mesophilic yeasts.

The highest UI, which was observed in obligate psychrophiles, was mainly due to the highest amounts of α-linolenic acid (C18:3, Δ9,12,15), at the expense of linoleic acid (C18:2, Δ9,12). At 4 °C, α-linolenic acid was found in all the obligate psychrophilic strains, accounting for up to 37% of the total FAs, whereas it was about 3.0% or even absent in the facultative ones. The remarkable amount of α-linolenic acid observed in obligate psychrophiles evokes an adaptive feature, which is probably due to the high activity of a Δ15 desaturase. α-Linolenic acid was generally absent in mesophilic yeasts, with the sole exception of K. marxianus L3.

Unexpectedly, facultative psychrophilic strains exhibited no significant increase of the UI at a lower temperature, in disagreement with previous works (McMurrough & Rose, 1973; Arthur & Watson, 1976; Watson & Arthur, 1976; Watson, 1978; Watson et al., 1978). This result raises the question of whether the increased unsaturation is really a general acclimatory response of yeasts to low temperature. Based on the results reported herein, the extent of unsaturation of FA could be considered one major adaptive feature exclusive of obligate psychrophilic yeasts.

A few recent studies reported that the response of diverse organisms to low temperatures includes the shortening of the FA chain (Chintalapati et al., 2004; Guschina & Harwood, 2006; Bahrndorff et al., 2007). The results reported herein confirm this trend only for mesophilic yeasts, which exhibited an increased concentration of C16:1 at the expense of C18:1 when cultured at 4 °C. On the contrary, the production of short-chain FA as a cold adaptation and/or an acclimation response in psychrophilic yeasts can be apparently excluded. The production of FA with the C18 chain length (in particular C18:1, C18:2, and C18:3) seems to be an adaptive feature of the yeasts from cold environments and may occur because elongation beyond C16 is necessary to introduce additional double bonds by Δ12 and Δ15 desaturases. The acclimation of the obligate psychrophilic strain R. glacialis DBVPG 4785 (selected as a representative strain) to different low temperatures is currently being investigated in greater detail (data not shown). This strain was capable of growing in the range between −3 and 15 °C with approximately the same YXS and YLX in the whole temperature range. As the temperature decreased, a progressive increase of both UI and FA with the C18 chain length was confirmed, mainly due to the increasing abundance of α-linolenic acid.

The present work improves information about the physiology of both facultative and obligate psychrophilic yeasts and may be useful for potential biotechnological application of these strains. Moreover, it sheds light on some physiological and ecological features, which are related to the growth temperature of psychrophiles, and suggests that obligate and facultative psychrophilic yeasts presumably use different metabolic strategies for adapting their life to different thermal conditions. Even if there is a continuum in temperature adaptation for life with wide or narrow growth temperature ranges, the results reported herein confirm that classification based on temperature limits was in good agreement with differences in the FA composition. The production of high amounts of PUFAs by obligate psychrophilic yeasts (in particular α-linolenic) correlated with adaptation to strictly cold environments. Over a time scale of many generations, the life of obligate psychrophilic yeasts was likely restricted to permanently cold environments, whereas facultative psychrophilic yeasts may have evolved in habitats subjected to wider temperature ranges.