The influence of Pseudomonas syringae on water freezing and ice melting

Abstract

Pseudomonas syringae is a widely spread plant pathogen known to have ice-nucleating proteins that serve as crystallization sites promoting ice growth at near-zero temperatures. Three temperatures that characterize water freezing and ice melting are (i) the freezing point of water, (ii) the temperature of coexistence of ice and water, and (iii) the melting point of ice. Here we show the influence of different concentrations of P. syringae on these three parameters. P. syringae appears to affect both the freezing point of water and the temperature of the coexistence of ice and water. Additionally, we propose a research technique for studying the freezing/melting process that is simple and requires no complex equipment.

Introduction

The ability of bacteria to initiate water freezing at near-zero temperatures was first discovered in 1974 using Pseudomonas syringae [1] and attributed to ice-nucleating proteins (INPs) attached to the bacterium’s outer membrane. INPs serve as nuclei of water crystallizing [2, 3]. The process of water freezing itself is quite complex, but its main aspects that are relevant to our work are discussed in article “Some peculiarities of water freezing at small sub-zero temperatures” [4].

The effect of different concentrations of P. syringae on water freezing has been the subject of many studies. Mostly, this process was controlled visually [5]. Aqueous or phosphate buffer was supplemented with a range of P. syringae concentrations, then droplets of the suspension were applied onto an aluminum plate, cooled from 0 to -10°C, and incubated for a few minutes, followed by visual counting of the frozen droplets. The temperature of solidifying of a certain number of droplets (e.g., 90%) was usually taken as a quantitative parameter of the process. This or similar methods are reported in [1, 6–8]. In [9], the same method was used to study the effect of lyophilized P. syringae cells on H₂O and D₂O freezing. In [10], individual droplets of a P. syringae solution were visually monitored and the freezing point of each droplet was recorded. In [11], freezing of P. syringae solutions was studied in microtubules under visual control. The above and other studies basically describe the temperature dependence of ice growth initiation at a certain concentration of living or lyophilized P. syringae.

The methods of visual droplet monitoring are clear but very labor-consuming and insufficiently accurate.

In our study, continuous temperature monitoring of a P. syringae suspension was used during both cooling and heating. This revealed the effect of different concentrations of P. syringae on three parameters: the freezing point of water (T_f), the temperature at which ice and water coexist in thermodynamic equilibrium (T_i-w), and the melting point of ice (T_m). As a control, the same experiments used different concentrations of E.coli cells.

Results

Fig 1 shows a typical temperature curve for a 0.5 ml water sample during freezing and melting. A test tube with room temperature buffer was placed in a thermostat pre-cooled to -11°C. A thermometer probe inside the test tube allowed continuous recording of sample temperature. As seen from Fig 1, for about 4 minutes the temperature decreases, with water remaining liquid and supercooled. At some freezing point T_f (in Fig 1, T_f = -9.7°C), the occurrence of an ice nucleator triggers water crystallization accompanied by heat release. After the 4th minute, the curve goes up dramatically reflecting the increasing temperature. From the 6th to the 9th minute, the sample is a mixture of ice and water that holds the temperature of coexistence of ice and water T_i-w (see the plateau in Fig 1, T_i-w = -0.11°C). With the entire water crystalized, the sample temperature decreases towards that of the thermostat (-11°C).

[Figure omitted. See PDF.]

A room-temperature sample (buffer, 20mM Tris-HCl, pH 7.5) was kept in a thermostat at -11°C for the time interval from zero to the 15th minute and after that transferred to a thermostat pre-heated to +25°C for further monitoring from the 15th minute onwards. The arrows show the curve stretches where the freezing point of water T_f, the temperature of coexistence of ice and water T_i-w, and the melting point of ice T_m were determined. For details of temperature determination, see Materials and Methods.

When transferred to a thermostat pre-heated to +25°C, the ice sample undergoes heating and melting at around the 16th-17th minute. In Fig 1, this process is seen as a bend in the curve which, ideally, should turn into a horizontal line whose position is determined by the ice melting point (T_m). In the case of pure water and an ideal device design, T_m and T_i-w must coincide (for details of temperature determination, see Materials and Methods).

Three temperatures that can be derived from the freezing/melting curve of the sample are important for understanding the activities of some proteins. For example, some antifreeze proteins or ice-binding proteins decrease the temperature of coexistence of ice and water T_i-w. In most publications, this phenomenon is called thermal hysteresis (see, e.g., [12–15]). Other studies show that some proteins can increase the melting point of ice T_m [16, 17]. Using the above method, here we show how different concentrations of P. syringae cells with ice-nucleating proteins attached to their outer membrane affect T_f, T_i-w, and T_m.

The effect of cells on water freezing and ice melting

Different concentrations of P. syringae cells were used in experiments on water freezing and ice melting, with E.coli cells as a control. Fig 2 shows typical freezing and melting curves for samples with different P. syringae concentrations. As seen, the presence of P. syringae triggers ice formation at higher temperatures, as compared to the buffer. E. coli cells do not affect water freezing (Fig 2C), that is, both with and without different E. coli concentrations, water freezing occurs at about -10°C. It should be noted that the freezing process of any solution is associated with the probability of occurrence of an ice nucleator, so the cooling/freezing curves are different even for identical solutions.

[Figure omitted. See PDF.]

Time dependence of sample temperatures during cooling (A and C) and heating (B and D). The curve for the 20mM Tris-HCl buffer, pH 7.5, is shown on all graphs as a black solid line. On A and B, the curves for P. syringae solutions with a cell concentration of 0.19 (--), 0.03(-·-), and 0.003(-··-) optical units are shown in red. On C and D, the curves for E coli solutions with a cell concentration of 0.02(--), 0.2(-·-), and 0.001(-··-) optical units are in blue. The inserts on A and C show the curve stretch from which the freezing point of water T_f has been derived.

In contrast, ice melting is characterized by good reproducibility. Fig 2B shows melting curves for P. syringae-containing ice. As seen, the curve shape is virtually unaffected by the difference in P. syringae concentration. Similarly, E. coli cells in any concentration do not affect the melting point of ice (Fig 2D).

The experiments with P. syringae and E. coli solutions allowed calculating T_f, T_i-w, and T_m as dependent on cell concentration in the samples. Fig 3 shows these dependencies.

[Figure omitted. See PDF.]

Dependence of T_f (A, A*), T_i-w (B, B*), and T_m (C, C*) on cell concentration. Figures marked with * are plotted on a logarithmic scale along the concentration axis. Red circles are for P. syringae solutions. Blue squares are for E. coli solutions. The empty circle shows T_f, T_i-w, or T_m for cell-free buffer averaged across ten replicates.

The data shown in Fig 3 allow the following conclusions:

1. None of the parameters T_f, T_i-w, and T_m depends on E. coli concentration in the solution.

2. Unlike T_m, the parameters T_f and T_i-w depend on P. syringae concentration in the solution.

Because freezing of supercooled water is a probabilistic process, the results presented in Fig 3A and 3A* are widely scattered, and a large variance is also observed for the buffer T_f data (see Materials and Methods). The probability of ice occurrence depends on the temperature of the solution and the number of ice nucleators. The growing P. syringae concentration increases the probability of water crystallizing due to INPs attached to the bacterium’s surface. Fig 3A* for P. syringae presents a linear dependence that allows calculating the relationship between T_f and the used cell concentrations from the following equation:(1)where C is cell concentration measured in optical units (o.u.). A large error (with an average deviation of ±1.5°C and maximal of 3.4°C) is explained by the randomness of ice nucleator appearance. This error and the exponential dependence of T_f on cell concentration (Fig 3A) allow the conclusion that at a P. syringae concentration above 0.15 o.u., T_f invariably falls within the range from 0°C to -2°C. Fig 3A clearly shows that at a cell concentration above 1.5 o.u., the temperatures vary from 0°C to -2°C. In other words, it is impractical to use P. syringae concentrations above 0.15 o.u. that corresponds to 1.2·10⁹ cell/ml.

Fig 3B and 3B* show the dependence of T_i-w on P. syringae concentration. A relatively small spread of data (± 0.03°C) results not only from the thermometer accuracy (± 0.01°C) but rather from a slight difference in ice growth and ice/water distribution observed in each individual test tube. Nevertheless, the dependence of T_i-w on P. syringae concentration is determined reliably with an average deviation of ±0.03°C and maximal of 0.05°C. What can this dependence be related to? Logically, it can be assumed that the cell surface-attached INPs display the ice-binding activity. If the ice surface is stabilized through these interactions, one can expect a shift in the ice melting/freezing equilibrium that will lead to an increase of T_i-w with the increasing number of cells in the solution. This conclusion might seem trivial if not for, e.g., ice-binding proteins whose effect is quite opposite: an increase in their concentration causes a lower T_i-w [12–15].

If P. syringae cells are assumed to be bound to the ice surface, T_m would also be expected to depend on their concentration. But Fig 3C shows that T_m is totally unaffected by either P. syringae or E. coli present in the solution. This result is surprising and difficult to explain. It is possible that the reason is the technical peculiarities of the experiment.

When ice grows in supercooled water, the test tube contains a mixture of small ice crystals and water. The thermometer probe is immersed in this mixture, so it measures the average T_i-w of the sample accurately enough.

Whereas in the melting experiment, the thermometer probe is surrounded by a large piece of ice whose melting begins from the outer surface close to the test tube walls that are in contact with the thermostat. Therefore, it is not the temperature of a mixture of water and ice that is measured but rather the temperature of the probe-surrounding ice. Perhaps this is the reason why we were unable to detect the dependence of T_m on the P. syringae concentration.

Conclusions

One of the results of this work is the proposed measurement technique. It turned out that using a solid-state thermostat maintaining a certain temperature and a high precision digital thermometer, it is possible to obtain well-reproducible data on peculiarities of water freezing and ice melting. This technique yielded Eq (1) linking the P. syringae concentration with the freezing point of a voluminous aqueous solution.

An important result is that this study has first revealed the dependence of T_i-w on the P. syringae concentration (Fig 3B), which is evidence for P. syringae binding to the ice surface. This result might seem trivial since surface INPs of P. syringae obviously must interact with the ice surface, but importantly, the obtained dependence means that only free-floating cells interact with ice, unlike those where ice has been already formed.

How to assess whether the interaction of cells with ice is strong or weak? In our experiments, P. syringae cells at a concentration of 0.3 o.u. (~2·10⁸ cell/ml) increased T_i-w by Δ = +0.13 ^оС. The equal T_i-w decrease (Δ = -0.13 ^оС) will be observed if the solution is supplemented, e.g., with 70 mM of a low molecular weight substance or 0.25% of ethyl alcohol [18, 19]. Another comparison example is ice-binding proteins which at high concentrations change T_i-w by more than 1^оС [20]. Thus, despite INPs, the P. syringae binding to the ice surface is weak.

Materials and methods

Cell culture experiments

E. coli BL21 (DE3) cells were grown on LB growth medium (VWR Life Science AMRESCO) at a temperature of 37°C. P. syringae cells (Pseudomonas syringae pv.syringae) grown on medium L (yeast extract 5.0 g/l; peptone 15.0 g/l; NaCl 5.0g/l) at a temperature of 26°C. All cells were grown in a liquid medium up to cell density OD600 = 1.0OU, then precipitated on a centrifuge at 6000g, washed twice with a solution of 20mM Tris-HCl, pH 7.5. The initial cell solution was diluted with a buffer solution of the same composition to the desired concentration. The concentration of cells was controlled by absorption at 600 nm.

Measurement of water freezing and ice melting

This study used a Biosan CH-100 (Latvia) solid-state thermostat with a fixed temperature of -11°C to measure water freezing or +25°C to measure ice melting. 0.5 ml samples in plastic 1.7 ml test tubes were placed in the thermostat. Temperatures were measured with a digital thermometer LT-300 (Russia) having an absolute accuracy of ±0.05°C and a relative accuracy of ± 0.01°C. The metal probe of the thermometer was fixed in a holder and positioned in the center of the tube. The temperature from the thermometer was recorded in real time on a computer. Fig 4 presents a schematic diagram of the described devise.

[Figure omitted. See PDF.]

Schematic diagram of a device for water freezing and ice melting studies.

Peculiarities of T_f, T_i-w, and T_m determination

From the graph of the temperature dependence of the sample on time (an example is shown in Fig 1) we determined three temperatures T_f, T_i-w and T_m.

The freezing point T_f was determined as the minimum value of the temperature at the time of the beginning of water crystallization in the sample.

The temperature of coexistence of ice and water Ti-w was defined as the temperature of the horizontal section of the curve after the moment of freezing of water. For example, in Fig 1, such a section of the curve is between the 6th and 10th minute.

To determine the melting temperature of ice T_m a graph of the first derivative of the temperature dependence curve was plotted. In Fig 5, the blue dotted line shows the section of the temperature dependence curve where ice melting occurs. The black solid line shows the graph of the first derivative of this section of the curve. The arrows in Fig 4 show how the temperature T_m is determined. Obviously, with such a definition of T_m, this parameter will always differ from the actual melting temperature of the T_mReal sample. If the sample had a sufficiently large volume and the thermometer could correctly measure its average temperature, then the curve would asymptotically tend to the T_mReal temperature. In Fig 4, the red dotted line shows the possible shape of the curve with a large sample volume. Moreover, a very accurate determination of the sample temperature is not possible due to the device. The temperature measurement takes place in the middle of the sample (ice), and its melting begins on the walls of the tube (which are in contact with the thermostat). Therefore, the temperature readings are always underestimated (in the case of melting) compared to the average temperature of the sample. Nevertheless, the temperature determination method shown in Fig 5 allows the T_m temperature to be determined in the same way for different samples and to compare the effect of different substances on the melting temperature of ice.

[Figure omitted. See PDF.]

The blue dashed line is an example of the time dependence of sample temperature during heating. The black solid line is the first derivative of the blue curve. The minimum of the first derivative graph allows determining the inflection point of the blue dashed curve (shown by the black arrow) from the position of which T_m is determined (shown by the blue arrow). The red dashed line shows a probable change in the experimental curve (blue) for the case when the sample is voluminous and the thermometer can precisely measure the average temperature of the sample. The red dashed line asymptotically tends to the real melting point of ice, T_mReal. The red arrow shows the real melting point of ice.

More than 10 measurement replicates of buffer (20mM Tris-HCl, pH 7.5) freezing and melting have been done to obtain the following three parameters: T_f = -9.8±0.5^оС, T_i-w = -0.12±0.01^оС, and T_m = -0.2±0.01^оС. Fig 1 shows only one curve as an example. Fig 6 shows three buffer freezing curves: one of them with a minimum deviation from the average T_f = -9.8°C and two others with a maximum deviation. The curves for buffer melting are not plotted here because they are identical to the curves shown in Fig 2B and 2C.

[Figure omitted. See PDF.]

For each curve, T_f values are indicated at the bottom of the graph. The T_i-w values are indicated at the top of the graph near the arrow. The average values of T_f and T_i-w across more than 10 replicates are -9.8 ±0.5^оС and -0.12±0.01^оС, respectively.

Acknowledgments

We are grateful to Alexei V. Finkelstein for fruitful discussions. We thank E.V. Serebrova for help in manuscript preparation.

Citation: Majorina MA, Veselova VR, Melnik BS (2022) The influence of Pseudomonas syringae on water freezing and ice melting. PLoS ONE 17(5): e0265683. https://doi.org/10.1371/journal.pone.0265683

About the Authors:

Maria A. Majorina

Roles: Data curation, Investigation, Validation

Affiliation: Institute of Protein Research, RAS, Pushchino, Moscow Region, Russia

Victoria R. Veselova

Roles: Investigation

Affiliation: Institute of Protein Research, RAS, Pushchino, Moscow Region, Russia

Bogdan S. Melnik

Roles: Conceptualization, Data curation, Methodology, Writing – original draft

E-mail: [email protected]

Affiliation: Institute of Protein Research, RAS, Pushchino, Moscow Region, Russia

https://orcid.org/0000-0002-9176-6452

References

1. Maki LR, Galyan EL, Chang-Chien M-M, Caldwell DR. Ice Nucleation Induced by Pseudomonas syringae. Appl Microbiol. 1974;28: 456–459. pmid:4371331

2. Gurian-Sherman D, Lindow SE. Bacterial ice nucleation: significance and molecular basis. FASEB J. 1993;7: 1338–1343. pmid:8224607

3. Wolber P, Warren G. Bacterial ice-nucleation proteins. Trends Biochem Sci. 1989;14: 179–182. pmid:2672438

4. Finkelstein A V. Some peculiarities of water freezing at small sub-zero temperatures. 2020. Available: http://arxiv.org/abs/2008.13682

5. Cochet N, Widehem P. Ice crystallization by Pseudomonas syringae. Appl Microbiol Biotechnol. 2000;54: 153–161. pmid:10968626

6. Goodnow RA, Harrison MD, Morris JD, Sweeting KB, Laduca RJ. Fate of Ice Nucleation-Active Pseudomonas syringae Strains in Alpine Soils and Waters and in Synthetic Snow Samples. Appl Environ Microbiol. 1990;56: 2223–2227. pmid:16348241

7. Kozloff LM, Schofield MA, Lute M. Ice nucleating activity of Pseudomonas syringae and Erwinia herbicola. J Bacteriol. 1983;153: 222–231. pmid:6848483

8. de Araujo GG, Rodrigues F, Gonçalves FLT, Galante D. Survival and ice nucleation activity of Pseudomonas syringae strains exposed to simulated high-altitude atmospheric conditions. Sci Rep. 2019;9: 7768. pmid:31123327

9. Weng L, Tessier SN, Smith K, Edd JF, Stott SL, Toner M. Bacterial Ice Nucleation in Monodisperse D 2 O and H 2 O-in-Oil Emulsions. Langmuir. 2016;32: 9229–9236. pmid:27495973

10. Pietsch RB, Vinatzer BA, Schmale DG. Diversity and Abundance of Ice Nucleating Strains of Pseudomonas syringae in a Freshwater Lake in Virginia, USA. Front Microbiol. 2017;8. pmid:28144237

11. Joly M, Attard E, Sancelme M, Deguillaume L, Guilbaud C, Morris CE, et al. Ice nucleation activity of bacteria isolated from cloud water. Atmos Environ. 2013;70: 392–400.

12. Kuiper MJ, Morton CJ, Abraham SE, Gray-Weale A. The biological function of an insect antifreeze protein simulated by molecular dynamics. Elife. 2015;4. pmid:25951514

13. Knight CA. Adding to the antifreeze agenda. Nature. 2000;406: 249–251. pmid:10917514

14. Devries AL. Glycoproteins as Biological Antifreeze Agents in Antarctic Fishes. Science (80-). 1971;172: 1152–1155. pmid:5574522

15. DeVries AL. Antifreeze glycopeptides and peptides: Interactions with ice and water. 1986. pp. 293–303.

16. Knight CA, DeVries AL. Melting Inhibition and Superheating of Ice by an Antifreeze Glycopeptide. Science (80-). 1989;245: 505–507. pmid:17750260

17. Celik Y, Graham LA, Mok Y-F, Bar M, Davies PL, Braslavsky I. Superheating of ice crystals in antifreeze protein solutions. Proc Natl Acad Sci. 2010;107: 5423–5428. pmid:20215465

18. Suzuki M, Ito K, Fushimi C, Kondo T. A Study of Cyclodextrin Complex Formation by a Freezing Point Depression Method. Chem Pharm Bull. 1993;41: 942–945.

19. Sweeney TE, Beuchat CA. Limitations of methods of osmometry: measuring the osmolality of biological fluids. Am J Physiol Integr Comp Physiol. 1993;264: R469–R480. pmid:8456999

20. Raymond JA, DeVries AL. Adsorption inhibition as a mechanism of freezing resistance in polar fishes. Proc Natl Acad Sci. 1977;74: 2589–2593. pmid:267952