Features

Infectious Causes

of Chronic Illness

Modeling the Earth

Microbiome

Cloudy with a Chance

of Microbes

ASM News

MicrobeFebruary 2012 Y The News Magazine of the American Society for Microbiology Y Vol. 7 Y No. 2

Cloudy with a Chance of MicrobesTerrestrial microbes swept into clouds can catalyze the freezingof water and may influence precipitation on a global scale

Brent C. Christner

In trying to disprove spontaneous gener-ation, Louis Pasteur helped to launchexperimental aeromicrobiology. Recog-nizing that virtually any microbe couldbe transported and disseminated in air,

he developed sampling procedures to study air-borne cells and soon showed that microbialabundances decreased at higher altitudes; pre-cipitation scrubbed dust and cells from the at-mosphere; and dust particles helped to dispersemicrobes in the atmosphere. Since then, aeromi-crobiology has contributed to our understand-ing of airborne pathogens and the life cycles ofother microbes.

Recent developments challenge the belief thatmicroorganisms are passive travelers in theatmosphere. Indeed, evidence that microor-ganisms affect meteorological processes isinvigorating research to define the role ofbiology in atmospheric processes and to un-derstand how those processes might affectthe abundance, dispersion, and viability ofaerosolized microbes. In particular, someairborne microbes, called ice nucleators(IN), efficiently catalyze ice formation. Thisice-nucleating phenotype (Ice�), first de-scribed nearly 40 years ago, appears impor-tant for dispersing microorganisms in pre-cipitation and might play a role in theformation of precipitation in clouds.

Independent Research Efforts Led to

the Discovery of Bacterial Ice

Nucleators

The formation of ice in clouds is importantfor the processes that lead to snow and mostrainfall. At temperatures warmer thanabout �36oC, aerosol particles are requiredeither to nucleate ice from water vapor or to

freeze water droplets directly. Abundant min-eral dusts and soot are presumed to serve as themajor sources of IN in clouds (Fig. 1). IN aresubstrates on which supercooled water mole-cules aggregate into a metastable form, called anembryo. As water molecules are added to bringan embryo to a critical size, the probability for aphase change in which ice crystals form steadilyincreases. The maximum temperature at whichIN can initiate the ice phase is specific to thenucleating material. Knowledge about the na-ture of IN in the atmosphere thus is fundamentalto understanding ice phase precipitation and theradiative balance on Earth.

In the late 1950s, atmospheric scientists be-

Summary

• Knowledge about microbial and other ice nucle-ators (IN) in the atmosphere is fundamental tounderstanding ice phase precipitation and theradiative balance on Earth.

• Independent research groups, one consisting ofatmospheric physicists and the other of micro-biologists, independently discovered ice nucle-ating bacteria during the 1970s.

• A number of bacteria are known to producesimilar ice-nucleating proteins that bind waterefficiently and may explain their superiority asIN over dust particles.

• While borne in clouds and dispersed widely byatmospheric currents, a large abundance of Ice�

microorganisms would be expected to influencefreezing and the processes which lead to precip-itation.

• Despite growing knowledge and interest, themicrobiota of the atmosphere remains a rela-tively unexplored frontier that appears ripe forfurther discovery.

Brent C. Christneris an Associate Pro-fessor in the De-partment of Biologi-cal Sciences,Louisiana State Uni-versity, BatonRouge; e-mail,xner@lsu.edu.

70 Y Microbe / Volume 7, Number 2, 2012

gan to explore the sources of warm-temperature IN in samples of snow,rain, and hailstones. Laboratory simu-lations of clouds indicate that super-cooling below �15oC is required for icenucleation in common mineral dustaerosols. Paradoxically, ice phase pre-cipitation occurs in clouds at tempera-tures warmer than �15oC.

Organic-rich soils are enriched in IN,according to Gabor Vali of the Univer-sity of Wyoming. Heating the soils to200oC abolishes this activity, accordingto his collaborator Russell Schnell. Indelving further during the summer of1970, Schnell left leaf and grass samplessoaking in distilled water and later no-ticed that these infusions were turbid.Testing indicated that the water froze at�1.3oC. “At that point, I suspected thatthe most active ice nuclei might be re-lated to living microorganisms,” helater commented. Shortly thereafter,University of Wyoming microbiologistLeroy Maki screened bacteria and fungifrom leaf surfaces for IN activity andfound a pair of isolates with ice nucle-ation temperatures identical to that of thedecaying leaf samples. By 1972, Makiidentified the isolate as a strain of Pseu-domonas syringae and demonstrated thatits ice-nucleating efficiency strongly de-pended on the culture incubation temper-ature and media nutrients.

Independently, Paul Hoppe at theUniversity of Wisconsin, who wasbreeding corn plants resistant to north-ern corn leaf blight, noticed in 1964 that frostdamage occurred only on plants inoculated withdried preparations of Exserohilum turcicum-infected leaves. His observation remained unex-plained until Steve Lindow joined Christen Up-per at the University of Wisconsin in 1973 andtested leaf extracts that became contaminatedby microbial growth. Plating the extract ledLindow to identify P. syringae as a strain thatactively incited freezing on plants. The specificmechanism by which the bacterium incited frostdamage remained elusive to Upper and Lindowuntil 1975, when the results from the group ofatmospheric scientists and microbiologists atthe University of Wyoming were brought totheir attention. Thus two research groups with

completely different scientific backgrounds andobjectives independently discovered ice nucleat-ing bacteria.

Advantages of Being an

Efficient Freeze Catalyst

In addition to P. syringae, proteinaceous IN arealso found in a variety of other bacteria such asP. viridiflava, P. fluorescens, Pantoea agglomer-ans, and Xanthomonas campestris and fungisuch as Fusarium avenaceum, as well algae,lichens, pollens, and freeze-tolerant animals.The ice nucleation protein of P. syringae (InaZ)is a 120- to 180-kDa protein containing a repet-

F I G U R E 1

A simplified overview of the processes required to initiate ice-phase precipitation inclouds.

Volume 7, Number 2, 2012 / Microbe Y 71

itive central domain flanked by nonrepetitive N-and C-terminal domains.

The ice-nucleating proteins from P. syringaeas well as other Ice� bacteria require a lipidenvironment for optimal activity. These pro-teins can form large homoaggregates, eachcontaining as many as 50 protein molecules,on the surface of the bacterial outer mem-brane. The tandem octapeptide repeats(AGYGSTET) of the central domain of theseproteins likely form a secondary structure ofantiparallel �-strands, creating a hydrogenbonding environment that favors the bindingof water molecules in a configuration similarto the ice crystal structure, according to mod-eling (Fig. 2).

The ability to bind water molecules in anordered fashion that mimics the ice lattice is acharacteristic that could explain the superiorityof biological IN over mineral dust and otherparticulates as warm-temperature IN. Nucleicacid sequence comparisons of ina genes fromvarious ice-nucleating bacteria indicate thatcoding portions of the N- and C-terminal geneproduct ends are homologous. Moreover, therepetitive sequences in the central domain varyfrom 832 to 1,280 amino acid residues. How-

ever, the number of repeats apparentlyis not a crucial determinant of ice-nu-cleating efficiency.

A considerable portion of microbeswith the Ice� phenotype associate withplants or are frank phytopathogens.Their ability to catalyze freezing at ele-vated temperatures may benefit thesemicroorganisms in one or more ways.For example, inducing plant tissues tofreeze would free nutrients from dam-aged cells and provide access to pro-mote disease. The Ice� phenotypemight also help microbes to survivefreeze-thaw cycling and could also pro-tect host tissues against damage fromsupercooling and low-temperature iceformation. Because water vapor tendsto accumulate on a frozen versus anunfrozen surface at subzero tempera-tures, Ice� species may survive longeron plants by preventing their desicca-tion during extended periods of low-temperature quiescence.

Moreover, the capacity of P. syringaeand other Ice� microorganisms to cata-

lyze the formation of ice could play a role in thedissemination of these microbes through theatmosphere. Due to their aerodynamic proper-ties, particles the size of bacterial cells can re-main suspended as aerosols for weeks. How-ever, the modeled atmospheric residence timefor an Ice� microbe is about 20-fold shorterthan that for cells incapable of ice nucleation.Thus, a mechanism that takes advantage of ice-phase scavenging might facilitate the return ofairborne microbes to surface habitats that sup-port their metabolism and growth.

Microbial Influences on

Meteorology and Precipitation

Microbes can be aerosolized from virtually anysurface and transported both horizontally andvertically in the atmosphere. They are ubiqui-tous in the near-surface and free troposphere,and are found in clouds at concentrations ofabout 104 cells m�3. There are even reports ofviable bacteria and fungi being collected from10- to 50-km altitudes in the stratosphere and50 to 100 km above the Earth in the meso-sphere. Nevertheless, very little is known about

F I G U R E 2

Space-filling model of the proposed three-dimensional structure of the ice nucleationprotein from P. syringae. Atoms in yellow and purple depict potential donors-acceptors ofhydrogen bonds in the ice lattice (left; cubic ice crystals) and protein (right). As shown,the ice lattice structure is directly superimposable on the protein surface structure, whichpossesses an ice-like binding site. Source: Steve Lindow and Andrey Kajava.

72 Y Microbe / Volume 7, Number 2, 2012

the flux, abundance, and diversity ofmicroorganisms in the Earth-atmo-sphere system.

Remarkably, certain atmosphericconditions may support microbialgrowth. Rates of heterotrophic produc-tion in supercooled cloud droplets sug-gest that cloud-borne bacterial biomasshas the potential to increase by as muchas 20% per day. Hence, microbes andtheir metabolic activities could affectmeteorological processes in the atmo-sphere both by changing cloud chemis-try and serving as nuclei for precipita-tion.

In 1978, David Sands at MontanaState University collected Ice� P. syrin-gae on petri plates containing selectivemedia that he thrust through the win-dow of a small airplane flying at alti-tudes up to 2,500 m. From such sam-pling experiments, he concluded that P.syringae is capable of disseminating inthe atmosphere via precipitation, per-haps over great distances. He specu-lated that airborne Ice� P. syringae andother bacterial IN trigger precipitationas a means of moving to new plant hostsvia the atmosphere, a process he calledbioprecipitation (Fig. 3). In support ofthis general concept, P. syringae de-taches from plant surfaces under dryconditions, becomes aerosolized, andcan be transported in the air both hori-zontally and vertically, with net up-ward fluxes on the order of 50 to 500CFU m�2 sec�1, according to Upper atthe University of Wisconsin and his col-laborators.

Soon after I met David Sands and hislongtime collaborator Cindy Morris in2005, I began collaborating with themin an effort to detect and quantify bio-logical IN in fresh snowfalls collected inthe vicinity of Bozeman, Mont. Wesoon expanded the range of our sam-pling to include sites in Louisiana, theFrench Alps and Pyrenees, Antarctica,and the Yukon. Although we did notexpect biological IN in every precipitation sam-ple analyzed, that is exactly what we found. Forinstance, melted snow and rain contain concen-trations of 5 to 500 IN per liter that are active at

temperatures warmer than �10oC. Based onthese analyses, about 95% of the IN are biolog-ical particles and at least 40% originate frombacteria. Further IN active above �10oC are up

F I G U R E 3

Proposed life cycle of P. syringae between the phyllosphere and the atmosphere. Cellsscrubbed in precipitation may encounter new plants directly or in rain or snow melt usedfor irrigation.

Volume 7, Number 2, 2012 / Microbe Y 73

to 8 times higher in the near-surface atmosphereof agricultural compared with forested areas,according to Noah Fierer and his colleagues atthe University of Colorado.

P. syringae is detected in freshly fallen precip-itation, atmospheric aerosols, and cloud water.To date, every strain of P. syringae that has beenisolated from snow and rain is Ice�, whereasthose isolated from other ecosystems it inhabitssuch as lakes, rivers, plants, and rock surfacesare less frequently ice nucleation active, accord-ing to Morris. This striking observation arguesstrongly that P. syringae has an active role as icenuclei in the atmosphere. However, P. syringaetypically constitutes only a small fraction of thetotal microflora deposited in the precipitation.In many precipitation samples that we haveanalyzed, we find significant levels of bacterialIN but no culturable populations of P. syringae,suggesting that P. syringae and comparable bac-teria maintain their ice nucleation activitieswhile losing culturability. Alternatively, suchsamples may contain novel microbial IN.

Effect of Biological IN on

Climate: a Game of Numbers

Biological IN particles might play a role in thehydrological cycle and radiative balance ofEarth. Enormous numbers—1024 to 1026 cells—of microorganisms inhabit leaf surfaces glob-ally. About one-third of the ice crystal residuesin clouds sampled over Wyoming are biologicalparticles, providing direct evidence for the in-volvement of bacteria, fungi, and/or plant mate-rial in ice-cloud processes, according to Kim-berly Prather of the University of California SanDiego and Paul DeMott at Colorado State Uni-versity. In addition, IN active above �25oC that

originate from the Amazon rainforest are dom-inated by biological particles, and the cloudcondensation nuclei are primarily composed ofsecondary organic aerosols from volatiles pro-duced by terrestrial biota, according to UlrichPoschl of the Max Planck Institute for Chemis-try and his colleagues. Taken with our findingsand those of Fierer and his colleagues, theseobservations support the hypothesis that plantcanopies are significant terrestrial sources ofbiological IN. Further, land use practices appearto influence the flux of biological IN to theatmosphere, and some microbial species dissem-inate in the atmosphere via precipitation.

Despite the fact that biologically derived ma-terials are the most active naturally occurringIN, there are few measurements of their concen-trations in the atmosphere. Moreover, their di-versity and sources have not been systematicallyinvestigated, and their role in generating precip-itation and influence over global climate re-mains speculative.

The challenges and complexities in measuringthe Earth-atmosphere system demand interdisci-plinary approaches to investigate the full meteoro-logical role of biological IN. This interdisciplinarystrategy is a fitting tribute to the atmospheric sci-entists, cloud physicists, microbiologists, andplant pathologists who crossed traditional bound-aries to discover microbial IN several decades ago.Thanks to them and subsequent research efforts,we now recognize that the atmosphere plays afundamental role in dispersing microorganisms,that clouds may support cellular reproduction,and that airborne microbes and their metabolicactivities could affect the global climate. Even so,the microbiota of the atmosphere remains a rela-tively unexplored frontier that appears ripe forfurther discovery.

ACKNOWLEDGMENTS

I gratefully acknowledge collaborators Cindy Morris and David Sands initially introducing me to the concept of bioprecipi-tation and providing a critical review of the text. Work on this topic has been partially supported by the Louisiana StateUniversity Office of Research and Economic Development, Louisiana Board of Regents, and National Aeronautics and SpaceAdministration (NNX10AN07A). This article was based on a keynote presentation during the 2011 ASM General Meeting inNew Orleans, La.

SUGGESTED READING

Amato, P., M. Menager , M. Sancelme, P. Laj, G. Mailhot, and A.-M. Delort. 2005. Microbial community in cloud water atthe Puy de Dome: implication for the chemistry of clouds. Atmospheric Environ. 39:4143–4153.Christner, B. C., C. E. Morris, C. M. Foreman, R. Cai, and D. C. Sands. 2008. Ubiquity of biological ice nucleators in snowfall.Science 319:1214.Constantinidou, H. A., S. S. Hirano, L. S. Baker, and C. D. Upper. 1990. Atmospheric dispersal of ice nucleation-activebacteria: the role of rain. Phytopathology 80:934–937.

74 Y Microbe / Volume 7, Number 2, 2012

DeMott, P. J, and A. J. Prenni. 2010. New directions: need for defining the numbers and sources of biological aerosols actingas ice nuclei. Atmospheric Environ. 44:1944–1945.Lee, R. E., G. J. Warren, and L. V. Gusta (ed.). 1995. Biological ice nucleation. APS Press, St. Paul, Minn., p. 370.Lundheim, R. 2002. Physiological and ecological significance of biological ice nucleators. Phil. Trans. R. Soc. London357:937–943.Morris, C. E., D. C. Sands, B. A. Vinatzer, C. Glaux, C. Guilbaud, A. Buffiere, S. Yan, H. Dominguez, and B. Thompson.2008. The life history of the plant pathogen Pseudomonas syringae is linked to the water cycle. ISME J. 2:321–334.Poschl, U., S. T. Martin, B. Sinha, Q. Chen, S. S. Gunthe, J. A. Huffman, S. Borrmann, D. K. Farmer, R. M. Garland, G. Helas,J. L. Jimenez, S. M. King, A. Manzi, E. Mikhailov, T. Pauliquevis, M. D. Petters, A. J. Prenni, P. Roldin, D. Rose, J. Schneider,H. Su, S. R. Zorn, P. Artaxo, and M. O. Andreae. 2010. Rainforest aerosols as biogenic nuclei of clouds and precipitation inthe Amazon. Science 329:1513–1516.Pratt, K. A., P. J. DeMott, J. R. French, Z. Wang, D. L. Westphal, A. J. Heymsfield, C. H. Twohy, A. J. Prenni, and K. A.Prather. 2009. In situ detection of biological particles in cloud ice-crystals. Nature Geosci. 2:398–401.Sattler, B., H. Puxbaum, and R. Psenner. 2001. Bacterial growth in supercooled cloud droplets. Geophysical Res. Lett.28:239–242.

Volume 7, Number 2, 2012 / Microbe Y 75