a novel group of archaea, extending the upper temperature limit of lifeto 113 8C. Extremophiles 1, 14–21

3 Brock, T.D. and Freeze, H. (1969) Thermus aquaticus gen. n. and sp. n.a non-sporulating extreme thermophile. J. Bacteriol. 98, 289–297

4 Fiala, G. and Stetter, K.O. (1986) Pyrococcus furiosus sp. nov.represents a novel genus of marine heterotrophic archaeabacteriagrowing optimally at 100 8C. Arch. Microbiol. 145, 56–61

5 Takai, K. et al. (2001) Distribution of archaea in a black smokerchimney structure. Appl. Environ. Microbiol. 67, 3618–3629

6 Schrenk, M.O. et al. (2003) Incidence and diversity of microorganismswithin the walls of an active deep-sea sulfide chimney. Appl. Environ.Microbiol. 69, 3580–3592

7 Robb, F.T. and Clark, D.S. (1999) Adaptation of proteins fromhyperthermophiles to high pressure and high temperature. J. Mol.Microbiol. Biotechnol. 1, 101–105

8 Daniel, R.M. and Cowan, D.A. (2000) Biomolecular stability and life athigh temperatures. Cell. Mol. Life Sci. 57, 250–264

9 Daniel, R.M. et al. (1996) The denaturation and degradation of stableenzymes at high temperatures. Biochem. J. 317, 1–11

10 Jaenicke, R. et al. (1996) Structure and stability of hyperstableproteins. Adv. Protein Chem. 48, 181–269

11 Adams, M.W.W. (1993) Enzymes and proteins from organisms thatgrow near and above 100 degrees C. Annu. Rev. Microbiol. 47, 627–658

12 Britton, K.L. et al. (1999) Structure determination of the glutamatedehydrogenase from the hyperthermophile Thermococcus litoralis andits comparison with that from Pyrococcus furiosus. J. Mol. Biol. 293,1121–1132

13 Vetriani, C. et al. (1998) Protein thermostability above 100 degrees C:a key role for ionic interactions. Proc. Natl. Acad. Sci. U. S. A. 95,12300–12305

14 White, R.H. (1984) Hydrolytic stability of biomolecules at hightemperatures and its implication for life at 250 degrees C. Nature310, 430–432

15 Hudson, R.C. et al. (1993) Glutamate dehydrogenase from theextremely thermophilic archaebacterial isolate AN1. Biochim. Bio-phys. Acta 1202, 244–250

16 Van der Casteele, M. et al. (1990) Pathways of arginine biosynthesis inextremely thermophilic archaeo- and eu-bacteria. J. Gen. Microbiol.

136, 1177–118317 Daniel, R.M. and Danson, M.J. (1995) Did primitive microorganisms

use non-heme iron proteins in place of NAD/P? J. Mol. Evol. 40,559–563

18 Pace, N.R. (1997) A molecular view of microbial diversity and thebiosphere. Science 276, 734–740

19 Vargas, M. et al. (1998) Microbiological evidence for Fe(III) reductionon early Earth. Nature 395, 65–67

20 Wachtershouser, G. (1998) A case for a hyperthermophilicchemolithoautotrophic origin of life in an iron-sulfur world. InThermophiles: The Keys to Molecular Evolution and the Origins ofLife? (Wiegel, J. and Adams, M.W.W., eds), pp. 47–58, Taylor andFrancis

21 Di Giulio, M. (2003) The ancestor of the Bacteria domain was ahyperthermophile. J. Theor. Biol. 224, 277–283

22 Stetter, K.O. (1998) Hyperthermophiles: Isolation classification andproperties. In Extremophiles: Microbial life in Extreme Environments(Horikoshi, K. and Grant, W.D., eds), pp. 1–24, Wiley-Liss

23 Sharp, R.J. et al. (1991) Heterotrophic thermophilic Bacilli. InThermophilic Bacteria (Kristjansson, J., ed.), pp. 19–50, CRCPress

24 Brock, T.D. et al. (1972) Sulfolobus: a new genus of sulfur oxidizingbacteria living at low pH and high temperature. Arch. Mikrobiol. 84,54–68

25 Zillig, W. et al. (1981) Thermoproteales: a novel type of extremelythermoacidophilic anaerobic archaebacteria isolated from Icelandicsolfataras. Zbl. Bakt. Hyg. I. Abt. Orig. C2, 205–227

26 Stetter, K.O. et al. (1983) Pyrodictium gen. nov., a new genus ofsubmarine disc-shaped sulfur reducing archaebacteria growingoptimally at 105 8C. Syst. Appl. Microbiol. 4, 535–551

0966-842X/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.doi:10.1016/j.tim.2003.12.002

|Research Focus Response

Response to Cowan: The upper temperature for life –where do we draw the line?

Kazem Kashefi

Department of Microbiology, University of Massachusetts, Amherst, MA 01003, USA

It was Louis Pasteur in the 19th century who made theconnection between spoiled food and microorganisms anddiscovered that heat treatment could be used to kill themicrobes that caused certain products to spoil. A series ofexperiments led the French microbiologist to determinethat temperatures in the order of 1108C destroyed allbacteria, therefore negating the popular theory of spon-taneous generation [1]. Pasteur also tested the heatresistance of fungal spores in a dry oven and found thata one-hour sterilization at 120–1258C was enough to killthe recalcitrant spores [1]. Sterilization standards were,therefore, not the arbitrary caprice of an impulsive mindbut the careful scientific product of one of the best inqui-sitive scientific minds in microbiology. In today’s auto-claves, a combination of pressure and steam to reachtemperatures in the order of 1218C has achieved what

Pasteur did more than a 100 years ago: to kill even themost resistant forms of life known to us.

Little could Pasteur predict that, more than a centurylater, a microorganism known as strain 121 would beisolated from a hydrothermal vent that could not only growat the standard 1218C temperatures of modern-day auto-claves, but could survive temperatures as high as 1308C[2], therefore raising all known standards for the highesttemperature for growth and survival of any living being.As stated in the article describing this discovery, ‘Theupper temperature limit for life is a key parameter fordelimiting when and where life might have evolved on ahot, early Earth; the depth to which life exists in theEarth’s subsurface; and the potential for life in hot,extraterrestrial environments.’ [2].

By growing at autoclave temperatures, strain 121raised the known upper temperature for growth of anyliving organism 88C higher than the previously recognizedCorresponding author: Kazem Kashefi (kkashefi@microbio.umass.edu).

Update TRENDS in Microbiology Vol.12 No.2 February 200460

www.sciencedirect.com

limit, held by Pyrolobus fumarii (1218C versus 1138C) [3],and also raised the survival temperature to 1308C forup to 2 hours, some 98C higher in temperature andover twice the period of time reported for P. fumarii(one hour at 1218C).

Although this is not a race to hold the record for‘hyperthermophilia’, comparisons between both micro-organisms are of the foremost importance to increase ourunderstanding on the evolution of life in hot environmentsand, possibly, on the potential of life existing in hot,extraterrestrial environments. However, these compari-sons have to be made on the basis of known facts and not onspeculations and wording that intends to downplay thesignificance of the reports, something that is apparent inthe first part of the article by D. Cowan. In this article, theauthor advises the reader to use caution when interpretingthe results reported for strain 121 [2]. In particular, hestates that ‘At first inspection, this organism would appearto have moved the bar up by some 88C…’ and arguesagainst the significance, and even the validity, of thisfinding, adducing reasons such as the difficulties inmeasuring growth rates in the presence of magnetiteand the similarities in optimum temperature for growth.Cell numbers are easily and accurately quantitated bydissolving the magnetite before cell counting; this can bedone by following a protocol [4] that was first described15 years ago [5] and has proven instrumental for accuratequantitation of cell numbers in a great number of studies.This protocol is clearly referenced by Kashefi and Lovley[2]. However, the author is right to advise the reader to usecaution when interpreting certain results; although bothmicroorganisms have comparable optimum temperaturesfor in vitro growth, the dramatic differences in uppertemperature limit for growth and survival and in physio-logy cannot be trivialized as they have profound impli-cations in understanding how these microorganismsimpact their ecosystems. Indicative of these substantialdifferences, strain 121 is a strict anaerobe that grows onpoorly crystalline Fe(III) oxide as the sole electron acceptor[2]. By contrast, P. fumarii is capable of microaerophilicgrowth, uses nitrate as an electron acceptor [2] and reportshave not been made as to whether this organism can growon any form of Fe(III). Corroborating the dramatic dif-ferences in their physiology, phylogenetic analyses of the16S rDNA and topR (reverse gyrase) genes placed strain121 closest to Pyrodictium occultum (96.0% similarity),closely followed by Pyrobaculum aerophilum (95.3%similarity) [2]. P. fumarii was not found to be among theclosest known relatives of strain 121. Therefore, on thebasis of phylogenetic and physiological studies, strain 121was found to represent a new genus in the Pyrodictiaceaefamily [2]. These data undermine any suggestion thatstrain 121 might be the same or similar to P. fumarii,as it is implied in Cowan’s article. Trivializing thedifferences in both their physiology and phylogeneticorigins is not only misleading, but also scientificallyincorrect. Paradoxically, there are more similaritiesbetween P. fumarii and P. aerophilum than there arebetween P. fumarii and strain 121, and even then itwould be wrong to suggest that P. fumarii andP. aerophilum are similar.

Caution also needs to be taken when referring topublished data without the context of the overall work.This is the case when the author concludes that a‘significant’ number of cells of P. fumarii surviveautoclaving. Blochl et al. [3] state that ‘An exponen-tially growing (middle log phase) culture and a culturein the stationary phase of isolate 1A were autoclaved inparallel for one hour at 1218C. Microscopic inspectionafterwards revealed that almost all of the stationaryphase cells were hyaline, while ,1% of the exponentialcells still exhibited a strong phase contrast inP. fumarii.’ Therefore, after only one hour of exposureto 1218C by autoclaving, all of the stationary phasecells and almost all (99%) of the exponentially growingcells were killed. The readers should note that thesenumbers were based on microscopic discriminationbetween ‘hyaline’ cells and cells exhibiting ‘a strongphase contrast’. It is then reasonable to speculate thata slightly longer incubation at 1218C might have killedall of the exponentially growing cells of P. fumarii.Moreover, Blochl et al. [3] also state, ‘Accordingly,when used as an inoculum (1% v/v) in fresh medium,autoclaved exponential phase cells grew up within threedays, while nonautoclaved controls grew up within oneday (data not shown).’ According to this, a 1% inoculumof a culture carrying an estimated 1% of survivorsreportedly grew when inoculated into fresh medium.The authors do not include data regarding growth yieldsor cell numbers, nor do they indicate if the growthtemperature used to revive the surviving cells was theoptimum (1068C) or the maximum (1138C) temperaturefor growth. It could not have been 1218C, otherwise allthe remaining cells would have been killed over theperiod of incubation (three days). And they conclude bysaying, ‘this indicates that a significant number of cellsin the exponentially growing culture survived autoclav-ing.’ Clearly, the use of the word ‘significant’ in this caseis misleading, because the data presented by Blochl et al.[3] showed that, at most, only 1% of viable cells mighthave survived autoclaving.

Despite the clear gap between growth and survivaltemperatures, and differences in both physiology andphylogeny, both microorganisms offer great opportunitiesto gain insight into the microbial processes that define hotenvironments. By increasing the upper temperature limitfor life, we have moved one step forward in this excitingfield and we can begin investigations aimed at under-standing the biological basis of growth and survival atthese temperature limits. Strain 121 has opened newhorizons to test new hypotheses on how life might havearose and evolved in a hot early Earth, which will probablyprovide invaluable tools for the exploration of microbiallife in hot extraterrestrial environments. This is not acompetition and should not be trivialized as such.

References

1 Geison, G.L. (1974) “Pasteur”. In Dictionary of Scientific Bibliography(vol. X) (Gillispie, C.C., ed.), pp. 350–416, Scribner

2 Kashefi, K. and Lovley, D.R. (2003) Extending the upper temperaturelimit for life. Science 301, 934

3 Blochl, E. et al. (1997) Pyrolobus fumarii, gen. and sp. nov., represents a

Update TRENDS in Microbiology Vol.12 No.2 February 2004 61

www.sciencedirect.com

novel group of archaea, extending the upper temperature limit of life to1138C. Extremophiles 1, 14–21

4 Kashefi, K. et al. (2002) Use of Fe(III) as an electron acceptor to recoverpreviously uncultured hyperthermophiles: isolation and characteriz-ation of Geothermobacterium ferrireducens gen. nov., sp. nov. Appl.Environ. Microbiol. 68, 1735–1742

5 Lovley, D.R. and Phillips, E.J.P. (1988) Novel mode of microbial energymetabolism: organic carbon oxidation coupled to dissimilatory reduc-tion of iron and manganese. Appl. Environ. Microbiol. 54, 1472–1480

0966-842X/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.doi:10.1016/j.tim.2003.12.006

|Microbial Genomics

The power in comparisons

William C. Nierman and Claire M. Fraser

The Institute for Genomic Research, 9712 Medical Center Drive, Rockville, MD 20850, USA

From the time that microbial genome sequencing andanalysis efforts began there has been a strong focus onunderstanding the biology of bacterial pathogens, thoseorganisms that share our planet’s biosphere to ourdetriment. This comes not only from a desire to improvepublic health through the development of novel prophy-laxes and therapeutics, but also from our innate desire tounderstand these dysfunctional (from our perspective)biological relationships. The ever-increasing number ofgenome sequences of pathogens has provided us withenhanced power to use comparative genomic analysis toexplore many of the aspects of the host–pathogenrelationship from the perspective of the pathogen. Theissues that can be addressed in this fashion includegenome evolution, genome determinants of host range andenvironmental survival, and the identification of crucialvirulence determinants.

The recent report of the sequence of three closely relatedBordetellas, known as Bordetella pertussis, Bordetellaparapertussis and Bordetella bronchiseptica, provides arich resource for addressing these issues [1]. B. pertussis isa human pathogen, the causative agent of whoopingcough. B. parapertussis infects both humans and sheep,causing whooping cough in human infants. B. bronchi-septica causes chronic respiratory infections in a range ofanimal hosts. The genome sizes and numbers of genesin these organisms increase in the following order:B. pertussis (4.1 Mb, 3816 genes), B. parapertussis(4.8 Mb, 4404 genes) and B. bronchiseptica (5.3 Mb, 5007genes). The comparison of the B. parapertussis chromo-some to that of B. bronchiseptica reveals large regions ofco-linearity with only eight rearranged blocks near thereplication terminus. This is contrasted by the largenumber (150) of rearrangements in B. pertussis relativeto B. bronchiseptica. In both of the smaller host-rangerestricted genomes that are rearranged relative toB. bronchiseptica, most of the rearrangements are boundedby insertion sequence elements (ISEs). Both of thesesmaller genomes contain very few unique genes relativeto B. bronchiseptica (114 for B. pertussis and 50 forB. parapertussis, excluding ISEs), and both the B. pertussisand B. parapertussis genomes are missing large segmentsof DNA that are present in B. bronchispetica. These

observations suggest that B. pertussis and B. parapertussisevolved from a common B. bronchiseptica-like ancestorand that adaptation to their more restricted host nicheswas accomplished largely through loss of functions, withvery little acquisition and/or expansion of new species-specific genes.

B. bronchiseptica is capable of environmental persist-ence,whereasB.pertussisandB.parapertussisareunabletopersist long outside the host. Genome comparisons provideinsight into this phenotypic difference. Most of the centraland intermediary metabolic pathways are conserved amongthe three. However, 1719 genes present in the B. bronchi-septica genome aremissing fromone or both of the other two.In addition, B. bronchiseptica is motile with a completeflagellar operon, whereas the flagellar operons in the othertwo species contain multiple frame-shifted and transposon-disrupted genes, explaining their non-motile phenotype.Although the B. bronchiseptica genome contains only 19pseudogenes, that of B. pertussis and B. parapertussiscontain 358 and 200 genes, respectively, which are inacti-vated by ISEs, frameshifts or in-frame stop codons. Many ofthese inactivated genes are involved in transport, smallmolecule metabolism, regulation and surface structures.Therefore, it appears that the inability of B. pertussis andB. parapertussis to persist in the environment is not aconsequence of differences in central metabolic pathwaysbut rather of the accumulation of losses of accessoryfunctions in the process of adaption to the restricted hostrange.

A comparison of genes that are well-documented to beinvolved in host interaction and pathogenicity providespossible insights into host range restrictions and possiblechallenges to assumptions about pathogenicity deter-minants. Species-specific variation in filamentous hemag-glutinins, the presence of a type-IV fimbrial system inB.bronchiseptica which is lacking in the other two, variationin the complement of autotransporters, different iron-uptake systems components, and the previously notedflagellar and motility differences could all contribute tohost range differences. Additional variation is observed inthe O-antigen genes and in the promoter regions of thepertussis toxin operons.

Type-III secretion systems and polysaccharide capsulesare often found to be crucial virulence determinants inorganisms that contain these systems [2,3]. Both of theseCorresponding author: Claire M. Fraser (cmfraser@tigr.org).

Update TRENDS in Microbiology Vol.12 No.2 February 200462

www.sciencedirect.com