Shedding Light on the Bioluminescence “Paradox” Although Luminescence Provides Host Squids with Obvious Advantages, How Does It Beneﬁt Light-Producing Bacteria?

Home , Bioluminescence, Bioluminescent bacteria

Shedding Light on the Bioluminescence “Paradox” Although luminescence provides host squids with obvious advantages, how does it beneﬁt light-producing bacteria?

Eric V. Stabb

he fascinating biochemistry, genetics, have special significance for this bacterium. Bi- and cell density-dependent regula- oluminescence offers many such puzzles, and it tion of bacterial bioluminescence is unlikely a single solution will solve them all. T provoke a challenging question. Nonetheless, it is an exciting moment in bio- What good is it to bioluminescent luminescence research, with recent advances of- bacteria? fering the promise of answering the longstand- There may be no single answer to this seem- ing question, “how can bioluminescence ingly simple question. However, two recent ad- help bacteria?” Although researchers learned vances shed new light on the problem. First, nearly a century ago that luminescence reduces studies of the symbiosis between the biolumines- oxygen and that symbiotic bacteria inhabit the cent bacterium Vibrio fischeri and the Hawaiian light organs of squids, what was unimaginable bobtail squid bring an ecologically relevant until very recently is our ability to analyze the V. niche for a bioluminescent bacterium into focus fischeri genome sequence and to combine this under the discerning lens of controlled labora- knowledge with the ability to genetically manip- tory experimentation. Second, progress in under- ulate these bacteria and observe them under standing the genetics of V. fischeri, including controlled laboratory conditions in the ecologi- genomic sequencing of a squid symbiont, is en- cally relevant environment of a natural squid abling researchers to analyze how luminescence host. integrates into the physiology of this bacterial species and to test specific hypotheses about Biochemistry and Genetics of what advantage light production confers on the V. fischeri Bioluminescence bacteria. In bacteria such as V. fischeri, light is generated Understanding how bioluminescence aids V. by an enzyme, luciferase, that contains two pro- fischeri in a squid light organ will likely not be teins, designated LuxA and LuxB (Fig. 1). the final word on how bioluminescence benefits LuxAB sequentially binds FMNH2,O2, and an bacteria. Differences among bioluminescent aliphatic aldehyde (RCHO) that are converted bacteria intimate that this ancient system con- to an aliphatic acid, FMN, and water. In turn, fers varied selective advantages. For example, they are released from the enzyme with the con- “cryptically luminescent” Vibrio pathogens, comitant production of light (Fig. 1A). Addi- such as Vibrio salmonicida, produce luciferase, tional proteins, LuxC, LuxD, and LuxE, are the enzyme responsible for bioluminescence, but responsible for (re)generating the aldehyde, little or no aldehyde substrate, and may use while another protein, LuxG, shuttles reducing Eric V. Stabb is an luciferase to form “dark reaction” hydrogen power from NAD(P)H to FMN to (re)generate Assistant Professor peroxide as a host-damaging virulence factor. FMNH2. in the Department In contrast, some strains of Vibrio logei pro- In V. fischeri, the luxC, luxD, luxE, and luxG of Microbiology at duce an accessory Y1 protein that shifts their genes flank luxA and luxB. These genes are the University of emitted light to a yellow wavelength, which may cotranscribed with luxI, while luxR is adjacent Georgia, Athens.

Volume 71, Number 5, 2005 / ASM News Y 223 FIGURE 1 ulated when intracellular AI concentra- tion exceeds a threshold. Thus, light is produced at high cell densities, such as A Biochemistry and physiology during growth in a squid light organ, Substratered Substrateox but not when planktonic cells are growing at low densities. O e– transport 2 NADH NAD+ H+ LuxG Bioluminescence at First Analysis + H motive force H2O Appears To Be a Drag on Cell Energy FMNH2 FMN Despite having a good understanding of ATP Products LuxAB the biochemical and genetic mechan- FMN ics of bioluminescence, researchers re- main uncertain over what its selective LuxAB LuxAB -FMNH2 advantage is to bacteria. In particular, H O O 2 the apparent costs in energy to cells 2 from bioluminescence make its exis- tence appear paradoxical. Light In addition to the sizable biosyn- O - 2 LuxAB -FMNH2 thetic cost in producing the Lux pro- LuxE + LuxC RCHO RCOOH teins, generating light consumes both biochemical reducing power and oxy- + NADP NADPH + ATP gen, seemingly competing for substrates LuxD + AMP + PPi with aerobic respiration, which recovers

O2- LuxAB -FMNH2 energy through electron transport (Fig. acyl-ACP 1A). Furthermore, energy stored as ATP RCHO is consumed in regenerating the aldehyde substrate (Fig. 1A). B Genetics and quorum sensing Indications that bioluminescence luxR luxI luxC luxD luxA luxB luxE luxG hinders cultured cells date at least as far back as E. Newton Harvey’s work at Princeton University in the early 1900s. Generate light Those findings were extended by J. Woodland (Woody) Hasting at Harvard OO University and his collaborators, who O showed that some relatively darker mu- N tants outgrew their brighter parents. Sim- H LuxR O LuxI ilarly, Paul Dunlap, now at the University AI-dependent Autoinducer (AI) of Michigan, characterized bright mu- transcriptional activator tants of V. fischeri that grew more slowly than do their relatively dim parent. V. fischeri bioluminescence. (A) Biochemistry and physiology. (B) Genetics and quorum In 1980, Ken Nealson and David Karl, sensing. currently at the University of Southern California and the University of Ha- waii, respectively, also found that cells to the other lux genes but not transcribed with expend appreciable energy for luminescence. them (Fig. 1B). Together LuxI and LuxR under- However, these two researchers did not find that lie a well-characterized quorum-sensing regula- bioluminescence slows growth rates. tory circuit, in which LuxI generates an autoin- These apparently inconsistent findings are dif- ducer (AI) that interacts with LuxR to stimulate ficult to interpret. Part of the explanation for lux transcription. The expression of lux is stim- the inconsistencies is that the best technology

224 Y ASM News / Volume 71, Number 5, 2005 Stabb’s Science: from Insects and Owls to Bioluminescent Bacteria and Squids Eric Stabb’s parents, both chemis- cards telling applicants when to E. scolopes hatchlings lack symbi- try teachers, encouraged his scien- be on hand for a phone interview,” onts but soon obtain V. fischeri tific interests from a very early age. he says. But the postcard arrived from their surroundings. Once When he became fascinated with too late for him to hold such an inoculated, an individual squid insects at age 8, his father made interview, scuttling his chances to carries V. fischeri cells in epitheli- him a butterfly net and mounting spend a summer doing field stud- um-lined crypts of a specialized board, and supplied him with glass ies in Alaska. organ. Light produced by those slides, pins, and, soon, his own In this case, zoology’s loss was bacterial cells helps the squid elude “knockout” jar plus a supply of car- microbiology’s gain. Stabb had predators, while the host squid bon tetrachloride. “They wouldn’t submitted what he calls a “fall- provides V. fischeri with nutri- let us use that stuff in college, and back” application to a program ents. I had a bottle at age 10,” Stabb sponsored by the National Sci- “I think it’s cool,” Stabb says, says. Soon those early insect-ori- ence Foundation to support un- referring to that symbiosis. “ I’m ented interests shifted, and he de- dergraduate research. He was ac- fascinated by bioluminescence and veloped a special liking for owls. cepted into this program, run by by the biology of the symbiosis.” He was so analytically minded as ASM-Carski Distinguished Teach- It also has a “gee whiz” appeal for a youngster that he collected owl ing Award recipient Ken Todar, students, he adds. “This is some- pellets, “looking for jawbones to and worked with Tim Donohue thing students can get fired up see what they had eaten,” he ad- studying photosynthetic bacteria. about, and it’s something they can mits. “I’m sure I sometimes left “I had no idea what microbiology learn experimental biology with. I them out, but I don’t remember was all about,” Stabb recalls. “I don’t know if my research will my mom ever complaining.” got hooked by what you could do have other practical benefits down Instead, his parents happily sup- experimentally with bacteria, es- the road—maybe, maybe not. But plied him with scientific kits, ma- pecially genetically.” Once hooked, I surely hope that some good sci- terials, and other support without he signed up to do more research, entists will get a start here.” being asked, he recalls. It was as if another course, and became “thor- Stabb, who grew up in Janes- “all my early grants were funded, oughly sold on prokaryotes.” ville, Wis., near the Rock River, is except most of the time I didn’t Stabb and his older brother, an avid runner. He competed in even apply for them,” he says. “It who holds a doctorate in applied high school and college, and still was like having a program man- physics and co-owns an engineer- trains for regular track sessions ager who anticipated what you ing consulting company, both with a group of running buddies. might want—and got it for you.” earned nearly straight A’s in their He and his wife Janice Flory, who Stabb, 37, now follows the con- high-school science courses, but ventional grant application pro- each with one exception. “He got is a project coordinator for the cess along with his teaching duties a B in physics, and I got one in Georgia Coastal Research Coun- in his professional role of assis- biology,” Stabb says. “There is cil, live in a neighborhood near tant professor in the department probably some deeper significance the UGA campus. “We both like of microbiology at the University to that, but I’m not sure what.” to garden, mostly flowers and of Georgia, Athens (UGA). His Stabb, who received both his ornamentals,” he says. “As introduction to microbiology B.S. and Ph.D. from the Univer- transplants from the north, we came pretty much by accident. Af- sity of Wisconsin, is now studying appreciate the longer growing ter he applied for a summer in- interactions between bacteria and season here.” ternship to do field biology in their hosts, specifically the light Alaska during his sophomore year organ symbiosis between the bac- Marlene Cimons at the University of Wisconsin terium Vibrio fischeri and the Ha- Marlene Cimons is a freelance writer (UW), Madison, “they sent post- waiian squid Euprymna scolopes. in Bethesda, Md.

Volume 71, Number 5, 2005 / ASM News Y 225 FIGURE 2 Squids Provide a Means for Scrutinizing Bioluminescence under Controlled Conditions A major advance for bioluminescence research came when Margaret McFall- Ngai and Edward Ruby, who now are at the University of Wisconsin, Madi- son, brought the V. fischeri-E. scolopes symbiosis into the laboratory. Al- though several other hosts of bioluminescent bacteria cannot be bred in Symbiotic bioluminescence. Light organs of E. scolopes juveniles under white light (left captivity, E. scolopes is an exception. panel) or lit from within by bioluminescent V. fischeri symbionts (right panel). It is thought In a series of ecological studies, Ruby that adult E. scolopes use this ventrally directed luminescence to obscure their and his collaborators found that V. fis- silhouette. Solid bars are ϳ100 ␮m. cheri in Hawaii are specifically adapted to E. scolopes, and they are more abun- of the day required comparisons of noniso- dant in waters inhabited by the host. genic strains, analyses of undefined pleiotro- Later, Karen Visick at Loyola University in pic mutants, or induction of luminescence by Chicago, Ill., provided evidence that bacteria AI, which also regulates non-lux genes. Re- derive an advantage from bioluminescence in cently, however, Grzegorz Wegrzyn’s group at this symbiosis. Specifically, although a luxA mu- the University of Gdansk in Poland used de- tant colonizes E. scolopes, this mutant (unlike fined strains to show that luminescence indeed its parental strain) does not persist well. One slows the growth of bacteria in cultures. Spe- possibility is that this mutant is attenuated be- cifically, he finds that a luxA mutant of V. har- cause of detrimental effects from expressing the veyi outcompetes its isogenic parent. Simi- full set of LuxCDBEG proteins in the absence of larly, we find that a luxCDABEG deletion a functional LuxA. However, we tested an in- frame luxCDABEG deletion mutant and came mutant of V. fischeri outcompetes its isogenic up with a result equivalent to Visick’s, namely wild-type parent in mixed culture when AI is poor persistence of this mutant in the squid. present. Thus, at least under some conditions, Thus, although bioluminescence slows V. fisch- light production slows bacterial growth. eri growth in culture, it enhances the bacteri- Presumably, however, luminescence confers um’s colonization of the squid light organ. How? an advantage to bacteria in some settings. For instance, bioluminescent bacteria naturally inhabit and illuminate the light-emitting organs of animals such as the Hawaiian bobtailed How Bacteria May Benefit squid, Euprymna scolopes (Fig. 2). Some biol- from Being Luminescent ogists argue that, in such associations, “what’s good for the host is good for the Researchers have proposed several hypotheses symbiont.” In this specific case, E. scolopes to explain how bioluminescence could confer apparently uses V. fischeri’s luminescence to advantages to light-producing bacteria (see ta- elude its predators, suggesting that increased ble). Several of these ideas downgrade the im- host fitness offsets the cost of bioluminescence portance of luminescence itself, describing it as to the bacteria, because the host “pays” the little more than an eye-catching distraction. bacteria back with nutrients. Rather, some researchers argue that the impor- Even so, without other constraints, dark mutant event occurs when luciferase consumes oxy- tants theoretically still should outcompete their gen or biochemical reducing power within the bright compatriots within a light organ, if forming bacterial cell. It may seem wasteful and thus only a minority population. However, available counterintuitive to burn reducing power and evidence indicates that such dark “cheaters” are oxygen without maximizing proton motive not found in symbiont populations, suggesting force and ATP generation; however, this is not luminescence yields additional advantages. unprecedented. Respiratory chains that are rel-

226 Y ASM News / Volume 71, Number 5, 2005 atively uncoupled from proton pumping help maintain redox balance, and in in- Table 1. How bioluminescence may benefit bacteria directly stances of “respiratory protection” they Possible Relevant consume oxygen rapidly enough to protect benefit process Mode of action oxygen-sensitive cytoplasmic proteins even Suffocate the O2 consumption Luminescence deprives nearby host epithelium when bacteria are in an aerobic environ- host of O2, attenuating reactive oxygen species ment. production This still leaves open the question of Holding their O2 consumption Luminescence lowers bacterial intracellular O2, breath protecting O2-sensitive enzymes and which is the more important reactant to increasing resistance to oxidative stress “burn,” oxygen or biochemical reducing Redox sink NADϩ Growth conditions lead to buildup of NADH, power? Many researchers believe the rele- regeneration and luminescence scavenges O2 to burn the vant reactant is oxygen, although for dif- excess reductant DNA repair Light production Luminescence stimulates light-dependent ferent reasons. For example, animal-asso- photolyase-mediated DNA repair in an ciated bioluminescent bacteria may otherwise dark environment consume oxygen to keep it away from the Avoid Light production Host distinguishes light-producing cells from sanctions dark ones and punishes the latter host. Depriving the nearby host epithelium of oxygen could attenuate the animal’s ability to produce antimicrobial oxygen radi- ensures that it receives light from the symbiosis. cals, or it could generate a low-oxygen environ- How animals could distinguish and selectively ment that facultative bacterial cells are better sanction dark bacterial cells that are mixed with suited to cope with than are obligately aerobic bright ones is not known, but finding crypto- host cells. On the other hand, bioluminescence chromes in the light organ merits further inves- may depress intracellular oxygen concentrations tigation. in the bacteria, either to increase resistance to Yet another proposal is that bacterial biolu- oxidative stress in general or to protect specific minescence stimulates DNA repair mediated by oxygen-sensitive enzymes. Proponents of these photolyase, an enzyme that uses visible-light explanations point to the relatively high affinity energy to fix pyrimidine dimers. According to of luciferase for oxygen as evidence for biolumi- Wegrzyn and his colleagues, when V. harveyi is nescence driving down oxygen levels. mutated by UV irradiation and then placed in Other scientists believe that the relevant re- the dark, the survival of lux mutants is attenu- actant is reducing power. The reducing power ated. Whether this treatment reflects an ecolog- for bioluminescence comes indirectly from the ically relevant condition is unresolved, but DNA NADH pool, and luciferase could therefore repair by luciferase and photolyase could be ϩ serve to recycle NAD cofactor. Jean-Jacques important even if something other than UV light Bourgois and his collaborators at the Universite´ damages the cellular DNA. Catholique de Louvain in Belgium recently bol- Wegrzyn and his collaborators also report a stered this argument with evidence that re- connection between the lux system and resis- ductant flows through luciferase only when tance to oxidative stress, although an alde- respiration is saturated. They speculate that lu- hyde-deficient dark mutant is also resistant. minescence becomes important for symbiotic Luciferase without aldehyde catalyzes a “dark bacteria when the other primary electron sink, reaction,” partially reducing oxygen to hydro- biomass production, becomes limited by spatial gen peroxide, which might prime oxidative constraints—for example, in the confines of a stress responses in this mutant. light organ. Despite such plausible benefits from burning oxygen or reducing power, the importance of Genetics and Genomics Shedding Light producing light itself cannot be dismissed. on Role of Bioluminescence in Bacteria Cheryl Whistler at the University of New Hamp- shire and Margaret McFall-Ngai propose that Genetic approaches should help us to evaluate cryptochrome photoreceptors allow squid to de- which among the proposed beneficial roles of tect bioluminescent symbionts and impose sanc- bioluminescence are valid. Bioluminescence tions on dark bacteria. This model has evolu- helps V. fischeri to colonize E. scolopes, and this tionary appeal in that it explains how the host symbiosis can be manipulated by establishing it

Volume 71, Number 5, 2005 / ASM News Y 227 FIGURE 3 be attenuated in their ability to colonize E. scolopes. Genome analyses reveal one photolyase homologue, and we are now generating a mutant to test its symbiotic phenotype. Although such ques- tions could be posed before genomic sequence information was available, genomics streamlines the process im- mensely. Analysis of the V. fischeri genome also allows us to broadly assess metabolic pathways and to determine how bioluminescence fits into an integrated physiological matrix. In particular, we can predict which pathways might com- plement luminescence under specific scenarios. For example, suppose that symbiotic V. fischeri are growing exclu- sively on peptides and building up excess NADH under low-oxygen conditions, depending on luminescence to scavenge remaining oxygen and regen- erate NADϩ. Under these conditions, how would V. fischeri generate energy? What other pathways would be regenerating NADϩ? If this “electron sink” model is correct, such pathways should be expressed during colonization of Enhanced bioluminescence in an arcA mutant of V. fischeri. Culture flasks of wild type the host, and a lux mutation would strain ES114 and its ⌬arcA mutant illuminated (top panel) or in the dark (bottom panel). make the cells more reliant on other NADϩ recycling pathways. Genomic analyses suggest that V. with genetically altered bacteria under con- fischeri would use certain amino acids to gener- trolled conditions. Such genetic manipulations ate ATP and acetate, and that cells could recover have become fairly standard, and include di- NADϩ by producing lactate or succinate. This rected mutations, expression studies using re- corresponds well with a 1942 report by Michael porters such as lacZ or green fluorescent pro- Doudoroff that 95% of the products of anaero- tein, and complementation analyses using a bic peptone catabolism by V. fischeri were ace- native V. fischeri plasmid. The above hypotheses tate, lactate, and succinate. If NADH buildup for the importance of luminescence provide nu- were problematic during growth on peptides, merous predictions that can be tested using such serine catabolism might be especially useful be- genetically engineered strains. cause serine deaminase does not reduce NADϩ Meanwhile, knowledge of the V. fischeri ge- in generating pyruvate. Genomic analysis nome sequence, which contains 4.3 million base suggests that the serine deaminase gene in pairs, is providing new ways to unravel the im- V. fischeri is coexpressed with a functional portance of bioluminescence in this bacterium. homologue of LuxG. The genome also re- First, it provides a ready means for identifying veals pathways of anaerobic respiration that and mutating specific genes to begin testing cur- would be important if appropriate terminal rent hypotheses about the metabolic effects of electron acceptors were present. All these luminescence. For example, if luminescence stim- genes and their regulatory sequences are avail- ulates photolyase-mediated DNA repair, and this able for mutational, reporter gene, or microar- reflects the symbiotic benefit of luminescence, ray analyses. then V. fischeri mutants lacking this gene should Genome analysis provides other useful and

228 Y ASM News / Volume 71, Number 5, 2005 unexpected information as well. For example, it al significance of luminescence. For instance, is often argued that luciferase has a higher affin- V. fischeri contains conserved regulatory path- ity for oxygen than respiration does, and there- ways that may modulate lux expression in re- fore might be useful in decreasing ambient O2. sponse to substrate availability. Among these, However this is based on assumptions that V. the ArcAB two-component regulatory cascade is

fischeri possesses a low-O2-affinity CyoABCD- especially intriguing. type terminal oxidase and on measurements of In Escherichia coli, ArcB senses redox state respiration in cells grown at relatively high O2, and, under reduced conditions, phosphorylates where a high-affinity respiratory system might ArcA, which acts as a DNA-binding transcrip- not be induced. Genomic analyses reveal that tional regulator. In V. fischeri ArcA and ArcB V. fischeri lacks CyoABCD but possesses a ho- may act as a regulatory switch if luminescence is molog of the high-O2-affinity CydAB-type ter- expressed to counteract oxidative stress or ex- minal oxidase and a terminal oxidase similar to cess reductants. Because ArcA is activated by phos- the very-high-O2-affinity FixN of Rhizobium phorylation under reduced conditions, ArcA-P loti, which has greater affinity for oxygen than may stimulate lux transcription if luciferase does luciferase. Such sequence similarities can- functions as an electron sink, or to repress lux if not tell us whether luminescence or respiration luciferase counteracts oxidative stress. Our pre- has higher O2 affinity, but they do suggest that it liminary data indicate that ArcA-P represses lux would be worthwhile to reexamine whether lu- in culture (Fig. 3) but that lux control by the minescence or respiration is more important for ArcAB system is derepressed in the host. If cor- setting ambient O2. rect, this would suggest that luciferase does not The genome sequence is also helpful in deter- benefit V. fischeri in the host squid by burning mining how the lux genes are regulated, which excess reductant, but may instead be expressed will likely provide clues regarding the function- to counteract oxidative stress.

ACKNOWLEDGMENTS I thank Jeffrey Bose for helping generate the figures, and Edward Ruby for many useful discussions. I was supported in this work by a CAREER award from the National Science Foundation (MCB-0347317), and by a Junior Faculty Research Grant from the University of Georgia Research Foundation. Genome information was provided by the Vibrio fischeri Genome Project, at http://ergo.integratedgenomics.com/Genomes/VFI, supported by the W. M. Keck Foundation. I wish to express my admiration and appreciation of the many scientists who have contributed to the field of bioluminescence, especially those who are not credited in this article or included in the Suggested Reading section below. E. Newton Harvey introduced the bibliography of his amazingly thorough book Living Light, published in 1940, by apologizing for its incompleteness; “The literature on luminescence is so vast that it is impossible to include (all) the important. . .papers.” The task has not gotten any easier in the intervening 65 years!

SUGGESTED READING Bourgois, J.-J., F. E. Sluse, F. Baguet, and J. Mallefet. 2001. Kinetics of light emission and oxygen consumption by bioluminescent bacteria. J. Bioengerg. Biomembr. 33:353–363. Czyz, A., K. Plata, and G. Wegrzyn. 2003. Stimulation of DNA repair as an evolutionary drive for bacterial bioluminescence. Luminescence 18:140–144. Greenberg, E. P. 1997. Quorum sensing in gram-negative bacteria. ASM News 63:371–377. Harvey, E. N. 1952. Bioluminescence. Academic Press, New York. McFall-Ngai, M. J. 1998. Pioneering the squid-vibrio model. ASM News 64:639–645. Nealson, K. H., and J. W. Hastings. 1979. Bacterial bioluminescence: its control and ecological significance. Microbiol. Rev. 43:496–518. Ruby, E. G. 1996. Lessons from a cooperative bacterial-animal association: The Vibrio fischeri-Euprymna scolopes light organ symbiosis. Annu. Rev. Microbiol. 50:591–624. Ruby, E. G., and M. J. McFall-Ngai. 1999. Oxygen-utilizing reactions and symbiotic colonization of the squid light organ by Vibrio fischeri. Trends Microbiol. 7:414–419. Visick, K. L., J. Foster, J. Doino, M. McFall-Ngai, and E. G. Ruby. 2000. Vibrio fischeri lux genes play an important role in colonization and development of the host light organ. J. Bacteriol. 182:4578–4586.

Volume 71, Number 5, 2005 / ASM News Y 229