eTh unGe a The Genu Serratia - Semantic Scholar · et al., 1980). The type strains of “Bacillus indicus” (Eisen-berg, 1886), “Erythrobacillus pyosepticus” (Fortineau, 1904),

CHAPTER 3.3.11The Genus Serratia

The Genus Serratia

FRANCINE GRIMONT AND PATRICK A. D. GRIMONT

The genus Serratia a member of the Enterobacteriaceae (seeIntroduction to the Family Enterobacteriaceae from thesecond edition.), is comprised of a group of bacteria that arerelated both phenotypically and by DNA sequence. The typespecies of the genus is Serratia marcescens. Some species andbiotypes of Serratia produce a nondiffusible red pigment,prodigiosin, or 2-methyl-3-amyl-6-methoxyprodigiosene(Williams and Qadri, 1980). The multiplication of red-pigmented Serratia was incriminated in the appearance ofbloodlike spots (e.g., on bread, consecrated wafers [sacra-mental Hosts], and polenta) with rather disastrous sociolog-ical consequences. In this context, several scholars havetraced the history of the genus Serratia back to antiquity(Gaughran, 1969; Harrison, 1924; Reid, 1936). However, sev-eral bacterial species outside the genus Serratia produce pro-digiosin or prodigiosin-like pigments (Williams and Qadri,1980) or many other kinds of red pigments, and the identityof microorganisms involved in these striking phenomena canonly be surmised.

Bizio (1823) named the red-pigmented microorganism heobserved on polenta Serratia marcescens. Ehrenberg (1848)named a motile bacterium isolated from red spots on food“Monas prodigiosa.” No cultures of these organisms were pre-served, but the name Serratia marcescens was preferred overthe name “Erythrobacillus pyosepticus”—a culture of whichwas preserved as ATCC 275 (Fortineau, 1904)—by Breed andBreed (1924, 1927) and by the editors of Bergey’s Manual ofDeterminative Bacteriology (Bergey et al., 1923). The nameSerratia marcescens is now universally accepted, and a neo-type strain has been designated (Martinec and Kocur, 1961a).

At the start of this century, more than 76 nomenspecieshad been described with red or pink pigmentation (Hefferan,1904), and 23 Serratia species were listed in the first editionof Bergey’s Manual (Bergey et al., 1923). This number pro-gressively decreased to five in the fifth edition of Bergey’sManual (Breed et al., 1957), and later to one species: S.marcescens (Ewing et al., 1959; Martinec and Kocur, 1960,1961a, 1961b, 1961c, 1961d). The only Serratia species recog-nized in the eighth edition of Bergey’s Manual was S. marce-scens (Sakazaki, 1974). Then, new objective approaches,such as numerical taxonomy and DNA relatedness appliedto strains recovered from diverse habitats, delineated anincreasing number of species in the genus Serratia. Five andseven species, respectively, were mentioned in the first edi-tion of The Prokaryotes (Grimont and Grimont, 1981) andin Bergey’s Manual of Systematic Bacteriology (Grimont andGrimont, 1984). Ten species are presently known to belongin the genus Serratia. These species (and synonyms) are:

1. Serratia marcescens Bizio 1823: “Typical” S. marcescens(Colwell and Mandel, 1965; Ewing et al., 1959; Martinec and

Kocur, 1961a); Serratia pattern 1 (Fulton et al., 1959); Serratiabiotype 1 (Bascomb et al., 1971); phenon A (Grimont andDulong de Rosnay, 1972; Grimont et al., 1977b); S. marce-scens DNA hybridization group (Steigerwalt et al., 1976). Theneotype strain is ATCC 13880, CDC 813-60, Grimont 504,and CCM 303 (Martinec and Kocur, 1961a). The nameappeared in the Approved Lists of Bacterial Names (Skermanet al., 1980). The type strains of “Bacillus indicus” (Eisen-berg, 1886), “Erythrobacillus pyosepticus” (Fortineau, 1904),“Bacillus sphingidis” (White, 1923a), and “Serratia anolium”(Duran-Reynals and Clausen, 1937) are all referable to thetaxonomic entity now known as S. marcescens. Strainslabeled “S. marcescens subsp. kiliensis” according to Ewinget al. 1962 were Voges-Proskauer-negative variants of S.marcescens.

2. Serratia liquefaciens (Grimes and Hennerty 1931)Bascomb et al. 1971: “Aerobacter liquefaciens” Grimes andHennerty 1931; “Aerobacter lipolyticus” Grimes 1961;Enterobacter liquefaciens (Grimes and Hennerty, 1931)Ewing 1963; phenon Clab (Grimont et al., 1977b); Serratialiquefaciens sensu stricto (Grimont et al., 1982a, 1982b). Thetype strain is ATCC 27592, CDC 1284-57, and Grimont 866.The name appeared in the Approved Lists of Bacterial Names(Skerman et al., 1980). It should be mentioned that the strainconsidered to be the type strain of “Aerobacter liquefaciens,”or of “Aerobacter lipolyticus” by Grimes (1961), was ATCC14460 (now the type strain of S. grimesii). Since this strainwas considered atypical, another strain (ATCC 27592) wasgiven as the type strain in the Approved Lists.

3. Serratia proteamaculans (Paine and Stansfield 1919)Grimont et al. 1978b: S. proteamaculans sensu stricto(Grimont et al., 1982a, 1982b). The type strain is ATCC19323, Grimont 3630, ICPB XP176, and NCPPB 245. Thename appeared in the Approved Lists of Bacterial Names(Skerman et al., 1980). S. proteamaculans and S. liquefacienswere thought to be synonymous on the basis of DNArelatedness (Grimont et al., 1978b). However, subsequentobservation of significant thermal instability of DNAhybrid fragments supported the separation of both species(Grimont et al., 1982a).

Biogroup RQ was named S. proteamaculans subspeciesquinovora (Grimont et al., 1982a, 1982b) with strain Grimont4364, CIP 8195, and ATCC 33765 as the type strain.

4. Serratia grimesii Grimont et al. 1982a, 1982b: phenonCld (Grimont et al., 1977b); S. liquefaciens hybridizationgroup (Steigerwalt et al., 1976). The type strain is ATCC14460 and Grimont 503. This strain had been considered asthe type strain of S. liquefaciens prior to the publication ofthe Approved Lists.

5. Serratia plymuthica (Lehmann and Neumann 1896)Breed et al. 1948: “Bacterium plymuthicum” Lehmann and

Prokaryotes (2006) 6:219–244DOI: 10.1007/0-387-30746-x_11

This chapter was taken unchanged from the second edition.

220 F. Grimont and P.A.D. Grimont CHAPTER 3.3.11

Neumann 1896; excluded from the genus Serratia (Ewing etal., 1959); “atypical” S. marcescens (Colwell and Mandel,1965); “S. marcescens var. kiliensis” according to Martinecand Kocur (1961d) (not Ewing et al., 1962); Serratia III (Man-del and Rownd, 1964); Serratia pattern 2 (Fulton et al., 1959);atypical S. rubidaea (Ewing et al., 1972, 1973); and phenonC2 (Grimont et al., 1977b). The type strain is Grimont 510,CCM 640, and ATCC 183. The type strains of “Bacteriumkiliense” reference is not an exact matchLehmann andNeumann 1986 and “Serratia esseyana” Combe 1933 are S.plymuthica.

6. Serratia rubidaeaStapp 1940 (Ewing et al., 1973): “Bac-terium rubidaeum” Stapp 1940; “Prodigiosus” VIII (Heffe-ran 1904); S. marinorubra ZoBell and Upham 1944; Serratiabiotype 2 (Bascomb et al., 1971); phenon B (Grimont andDulong de Rosnay, 1972; Grimont et al., 1977b); S. rubidaeaDNA hybridization group (Steigerwalt et al., 1976). Thename appeared in the Approved Lists of Bacterial Names(Skerman et al., 1980). The type strain (neotype) is ATCC27593, CDC 2199–72, and Grimont 864. The former typestrain (holotype) of S. marinorubra was NCTC 10912, ATCC27614, and Grimont 288. However, the Approved Lists ofBacterial Names (Skerman et al., 1980) listed S. marinorubrawith ATCC 27593 as the type strain. Thus, both names (S.rubidaea and S. marinorubra) were made objective synonymsand the name S. rubidaea has priority. Three subspecies canbe delineated: S. rubidaea subsp. rubidaea S. rubidaea subsp.burdigalensis, and S. rubidaea subsp. colindalensis (Grimontet al., manuscript in preparation).

7. Serratia odorifera Grimont et al. 1978a: Strains similarto the unclustered Serratia strain 38 (Grimont and Dulongde Rosnay, 1972; Grimont et al., 1977b). The type strain isATCC 33077, CDC 1979-77, Grimont 1073, ICPB 3995, andNCTC 11214. The name appeared in the Approved Lists ofBacterial Names (Skerman et al., 1980).

8. Serratia ficaria Grimont et al. 1979c: the type strain isATCC 33105, Grimont 4024, CIP 79.23, and ICPB 4050.

9. Serratia entomophila Grimont et al. 1988: the type strainis ATCC 43705, Jackson A1, and CIP 102919.

10. Serratia fonticola Gavini et al. 1979: Lysine-positiveCitrobacter-like (Steigerwalt et al., 1976). The type strain isATCC 29844, CUETM 77.165, Grimont 4011, and CIP 78.64.The name appeared in the Approved Lists of Bacterial Names(Skerman et al., 1980).

The inclusion of the genus Serratia in the tribe Klebsielleaeis no longer tenable. Studies on DNA relatedness, immuno-logical cross-reaction between isofunctional enzymes, andthe physical properties, regulation, and amino acid sequencesof enzymes have all shown the genus Serratia to be consis-tently different from the group composed of the generaEscherichia, Shigella, Salmonella, Citrobacter, Klebsiella, andEnterobacter (reviewed by Grimont and Grimont, 1978a).

Until the late 1950s, Serratia spp. were rarely isolated fromhuman patients. Later on S. marcescens became more andmore frequently involved in nosocomial infections and non-pigmented S. marcescens strains are now a serious threatin surgical and intensive care units (Daschner, 1980; vonGraevenitz, 1980; Yu, 1979).

Ecology

For a long time, the confused status of Serratiataxonomy prevented any precise knowledge ofthe habitat of Serratia species. After the afore-mentioned Serratia species had been defined,however, it became apparent that they occupieddifferent habitats. Table 1 shows the distributionof the major species of the Serratia strains iso-lated from small mammals and their territories,and from water, plants, and hospitalized humanpatients. Table 2 shows the distribution of themajor biotypes of the S. marcescens strains iso-lated from these same habitats.

Serratia in Water and SoilWater is probably the principal habitat of S.plymuthica. Nomenspecies resembling S.plymuthica (Breunig’s Kiel bacillus, “Bacillusminiaceus,” “Serratia miquelii,” and “S. essey-ana”) have also been isolated from water. Sea-water isolates belong to the same species asterrestrial water isolates. “Serratia marinorubra”

Table 1. Percent distribution among Serratia species of 1,543 strains isolated from different habitats.

aRodents and shrews.bFrom plants and soil around the traps where small mammals were captured.cAll plants except figs and coconuts.

Species

Percentage of isolates in habitat

Small mammalsa

(92 isolates)Animal territoriesb

(51 isolates)Water

(155 isolates)Plantsc

(137 isolates)Hospitalized patients

(1,108 isolates)

S. marcescens 10 2 75 10 97S. plymuthica 4 2 1 5 0S. liquefaciens 43 55 11 38 2S. proteamaculans 39 29 8 32 0S. grimesii 3 10 5 7 0.5S. rubidaea 0 0 0 1 0.2S. odorifera 0 0 0 4 0.1S. ficaria 0 2 0 4 0Total percentage 100 100 100 100 100

CHAPTER 3.3.11 The Genus Serratia 221

was not a true marine species, as it showedno sodium requirement (P. Baumann, personalcommunication). Many strains of S. fonticolahave been isolated from well waters and springs(Gavini et al., 1979). In a systematic search forSerratia in river water, 150 strains were isolatedand were distributed in the following species:S. marcescens (75%), S. liquefaciens (11%), S.proteamaculans (8%), S. grimesii (5%), and S.plymuthica (1%) (F. Agbalika, F. Grimont, andP.A.D. Grimont, unpublished observations). Theselective medium used (caprylate-thallous agar)did not allow isolation of S. fonticola.

Strains producing the non-diffusible red pig-ment prodigiosin seem to be toxic to protozoa(Groscop and Brent, 1964), and this may be anecological advantage in water and soil. However,it seems that pigmented bacteria are more oftenisolated from unpolluted water (from springs orwells) than from polluted water (river waterdownstream from cities).

In soil, S. marcescens might play a role in thebiological cycle of metals by mineralizing organiciron and dissolving gold and copper (Parès,1964). A mineralization role has also been attrib-uted to cold-tolerant Serratia associated withlow-moor peat (Janota-Bassalik, 1963).

Serratia on PlantsSerratia proteamaculans was once isolated froma leaf spot disease of the tropical plant Proteacynaroides (the King Protea) (Paine and Stans-field, 1919). However, the experimental lesionscaused by S. proteamaculans on detached leavesof Protea suggest a hypersensitivity reaction(Paine and Berridge, 1921). Similar lesions wereobtained with other species of Serratia (Grimontet al., 1978b). Inoculation of S. marcescens ICPB2875 and S. marinorubra ICPB 2881 on tobaccoand bean leaves also produced a typical hyper-sensitivity reaction (Lakso and Starr, 1970). Aroot disease complex of alfalfa (Medicagosativa) involving a Fusarium sp., a Pseudomonas

sp., and a bacterium named “Erwinia amylovoravar. alfalfae” was observed by Shinde andLukezic (1974). These isolates of “Erwiniaamylovora var. alfalfae” were later identified asSerratia marcescens biotype A4a (Grimont et al.,1981).

Characteristic Serratia populations were foundin figs. S. ficaria was recovered from most figscollected in California, Tunisia, France (Grimontet al., 1979c), Sicily (Giammanco and Amato,1982; Grimont and Deval, 1982), and Greece(P.A.D. Grimont, unpublished observations).This species could be associated with S. marce-scens (biogroups A1, A2, or A3) or other genera.Figs of the Calimyrna variety (Smyrna varietyadapted to California) deserve special mention.Calimyrna figs contain only pistillate flowers andin order to become edible, they need to be pol-linated specifically by the fig wasp Blastophagapsenes. The fig wasp has a life cycle limited to thecaprifig, an inedible fruit produced by the capri-fig tree. When a young female fig wasp makes herway out of a caprifig, she covers herself withpollen and a specific fungal and microbial flora(Phaff and Miller, 1961). The fig wasp will thenenter an unpollenated caprifig of a new crop,pollinate it, oviposit in a pistillate flower, and die;or it may be carried by the wind to a Calimyrnafig—a cul-de-sac in the wasp life cycle—wherethe fig wasp will deposit pollen and the micro-bial/fungal flora in its desperate attempt to ovi-posit. Since 1927 (Caldis, 1927; Phaff and Miller,1961; Smith and Hansen, 1931), a red-pigmentedand a nonpigmented Serratia have been repeat-edly isolated from caprifigs, pollinated Cal-imyrna figs, and fig wasps. The pigmented andnonpigmented isolates were identified as Serratiamarcescens biotype A1b and S. ficaria, respec-tively (Grimont et al., 1979c; Grimont et al.,1981). The fact that the Calimyrna fig is inter-nally sterile until the fig wasp enters it providesan ecological niche for these Serratia strains. Themultiplication of Serratia in caprifigs or Cal-imyrna figs is limited, and these organisms do not

Table 2. Percent distribution among Serratia marcescens biogroups of 1,210 strains isolated from four different habitats.

aIsolated at the Pellegrin Hospital (Bordeaux, France) from 1968 through 1975.bIn addition, 72 strains of biotype A2b were isolated from infants (mostly from feces) in the neonatology ward.

Biogroup

Percentage of isolates in habitat

Water (107 isolates) Plants and rodents (37 isolates) Hospitalized patients (1,066 isolates)a

A1 12 19 0.1A2/6 23 5 7b

A3 21 38 7A4 34 38 26A5/8 7 0 47TCT 3 0 13Total 100 100 100


cause fig spoilage. Washing of a syconium cavitycan yield 10 to 300 colony-forming-units (CFU)of S. ficaria.

Serratia rubidaea has been repeatedly isolatedfrom coconuts bought in France (originatingmostly from Ivory Coast) and in California (Gri-mont et al., 1981). The three subspecies (formerbiotypes B1 to B3) have been associated withcoconuts, and when present, S. rubidaeanumbered 3! 104 to 6! 106 CFU per gram ofcoconut milk or flesh (Bollet et al., 1989). It isbelieved that coconut palms have been disse-minated worldwide from an Indo-Pacific origin.It would thus be interesting to compare thebacterial flora of coconuts collected in diversecountries.

In the course of an ecological survey and usingthe selective caprylate-thallous agar medium,Serratia species were frequently found associatedwith plants (Grimont et al., 1981). With respectsto the presence of Serratia, plants (other than figsand coconuts) were classified into three preva-lence groups. In the first group, which is com-posed of vegetables, mushrooms, mosses, anddecaying plant material (i.e., wet plants), about54% of the samples carried Serratia in countsvarying from 2,000 to over 20,000 CFU per gramof plant (highest counts in mushrooms andleaves of radish, lettuce, cauliflower, and brusselssprout). In the second group, which is composedof grasses, Serratia prevalence was about 24%. Inthe third group composed of trees and shrubs,8% of the samples (leaves) contained Serratia(10 to 300 CFU per gram).

Serratia strains isolated from plants other thanfigs and coconuts were distributed among eightSerratia species, although S. liquefaciens and S.proteamaculans were predominant (Table 1).The selective medium used (caprylate-thallousagar) was unable to support the growth of S.fonticola. S. entomophila was not isolatedalthough this species can grow on the selectivemedium. Pigmented S. marcescens biotypes wererarely isolated from plants. Nonpigmented S.marcescens biogroups A3 and A4 were isolatedfrom plants, but biogroups A5/8 and TCT werenot (Table 2).

Vegetables used in salads might bring Serratiastrains to hospitals and contaminate the patient’sdigestive tract. S. marcescens, S. liquefaciens, andS. rubidaea were found in 29%, 28%, and 11%(respectively) of vegetable salads served in ahospital in Pittsburgh (Wright et al., 1976). Sim-ilar results were found in a Paris hospital(Loiseau-Marolleau and Laforest, 1976). How-ever, it still needs to be proven that biotypes/serotypes found in patients are the same as thosefound in salads. In the first described case of S.ficaria infection, the patient was very fond of figs(Gill et al., 1981).

Serratia in Insects

There is an extensive literature on Serratia asso-ciated with insects. This topic has been reviewedin detail (Bucher, 1963a; Grimont and Grimont,1978a; Steinhaus, 1959). The insects involvedbelong to numerous species and genera of theorders Orthoptera (crickets and grasshoppers),Isoptera (termites), Coleoptera (beetles andweevils), Lepidoptera (moths), Hymenoptera(bees and wasps), and Diptera (flies). Taxonomicuncertainties make a retrospective evaluation ofthe role of Serratia spp. in insect infections diffi-cult. Red-pigmented Serratia were easily recog-nized (although not determined as to specieswith any certainty), whereas nonpigmentedstrains were often referred to genera other thanSerratia. For example, S. proteamaculans or S.liquefaciens strains were named “Bacillus noctu-arum” (White, 1923b), “Bacillus melolonthae liq-uefaciens” (Paillot, 1916), “Paracolobactrumrhyncoli” (Pesson et al., 1955), and “Cloaca” Btype 71–12A (Bucher and Stephens, 1959); non-pigmented S. marcescens strains were named“Bacillus sphingidis” (White, 1923a) and “Bacil-lus apisepticus” (Burnside, 1928).

The distribution among the various Serratiaspecies of 48 collection strains isolated frominsects (Grimont et al., 1979b) showed a pre-dominance of S. marcescens and S. liquefaciens.The absence of S. plymuthica (a pigmented spe-cies) in this collection is unexplained. Steinhaus(1941) reported the presence of S. plymuthica inthe gut of healthy crickets (Neombius fasciatus),but the taxonomic schemes in vogue at that timedid not allow a definite identification of S.plymuthica. The rarity of S. rubidaea (also a pig-mented species) in insects might be explained byits inability to produce chitinase—a virulencefactor for insect-associated Serratia species(Lysenko, 1976).

The red-pigmented Serratia isolates associatedwith the fig wasp Blastophaga psenes and for-merly identified as S. plymuthica (Phaff andMiller, 1961) were reidentified as S. marcescensbiotype A1b (Grimont et al., 1981). It is notewor-thy that S. ficaria was isolated from both the figwasp Blastophaga psenes and from a black ant,i.e., Hymenoptera (Grimont et al., 1979c).

S. liquefaciens and S. marcescens were foundin sugar-beet root-maggot development stages(Tetanops myopaeformis), suggesting an insect-microbe symbiosis, as well as a nutritional inter-dependence (Iverson et al. 1984). A relationshipappeared to exist between adult fly emergenceand enzymatic chitin degradation of the pupar-ium by the bacterial symbionts.

S. marcescens biotype A4b was repeatedly iso-lated from diseased honeybee (Apis mellifera)larvae in Sudan (El Sanoussi et al., 1987).


Serratia marcescens, S. proteamaculans andS. liquefaciens are considered potential insectpathogens (Bucher, 1960). They cause a lethalsepticemia after penetration into the hemocoel.More than 70 species of insects were found to besusceptible to inoculation with Serratia (Bucher,1963a). The lethal dose (LD50) of inoculated Ser-ratia (by intrahemocoelic injection) was calcu-lated for several insects 10–50 Serratia cells pergrasshopper (Bucher, 1959), 5.1 cells per adultbollweevil (Slatten and Larson, 1967), 7.5 and14.5 cells per third and fourth instar larva(respectively) of Lymantria dispar (Podgwaiteand Cosenza, 1976), and 40 cells per Galleriamellonella larva (Stephens, 1959). The LD50 ofingested S. marcescens is much higher. Thehemolymph of insects—normally bactericidal fornonpathogens—cannot prevent multiplication ofpotential pathogens (Stephens, 1963). Lecithi-nase, proteinase, and chitinase play a role in thevirulence of Serratia for insects, and purifiedSerratia proteinase or chitinase is very toxicwhen injected into the hemocoel (Kaska, 1976;Lysenko, 1976). Serratia strains in the insectdigestive tract probably originate from plants.The multiplication of Serratia strains in the insectdigestive tract has not been quantitatively stud-ied. Antibacterial substances in ingested leavesmight interfere with bacterial multiplication, butSerratia strains were found resistant to these(Kushner and Harvey, 1962). How potentialpathogens such as Serratia can enter thehemolymph from the gut is generally unknown.However, spontaneous gut rupture, which hap-pens in about 10% of grasshoppers, may allowSerratia strains to invade the hemocoel (Bucher,1959). Direct injection occurs when Itoplectisconquisitor contaminated with Serratia stingshost pupae to oviposit into their bodies (Bucher,1963b). Serratia epizootics are common amongreared insects (Bucher, 1963a). However, untilrecently, no genuine epizootic of Serratia infec-tion among insects had been observed in the field.

Strains of S. entomophila and S. proteamacu-lans can be pathogenic for the grass grub Coste-lytra zealandica, which is a major pasture pest inNew Zealand (Stucki and Jackson, 1984; Troughtet al., 1982). Larvae of Costelytra zealandica feedon grass, clover, and other plant roots. Typically,their populations grow to a peak (about 600 lar-vae/m2) in 4 to 6 years after the pasture is sownand then collapse (to about 50 larvae/m2). Grassgrub population collapse was found associatedwith the presence of a disease called amber dis-ease (Trought et al., 1982). S. entomophila and S.proteamaculans were isolated from naturallyinfected larvae and shown experimentally to pro-duce the disease when transmitted orally tohealthy larvae (Grimont et al., 1988; Stucki andJackson, 1984). The bacteria are ingested from

the soil. Infected larvae stop feeding within a fewdays, become translucent and then amber col-ored and lose weight until death occurs 4 to 6weeks later. The bacteria colonize the gut andcause disease symptoms without invading thehemocoele. Field trials have shown that controlof the grass grub was feasible by application ofS. entomophila suspensions on pastures (Jacksonand Pearson 1986). Reductions of 30 to 59% inthe larval populations were obtained, with 47%of the remaining larvae being infected. Such bac-terial treatment resulted in a 30% increase ingrass production (dry matter).

Serratia in VertebratesSerratia has been associated with chronicinfections of cold-blooded vertebrates: nodularinfection of Anolis equestris, the Cuban lizard(Duran-Reynals and Clausen, 1937); subcutane-ous abscess of iguanid lizards (Boam et al., 1970);arthritis in the lizard Tupinambis tequixin (Ack-erman et al., 1971); and ulcerative disease in thepainted turtle Chrysemys picta (Jackson and Ful-ton, 1976). Serratia strains have also been recov-ered from the healthy, small, green pet turtlePseudemys scripta elegans (McCoy and Seidler,1973) and from geckos and turtles in Vietnam(Capponi et al., 1956).

Poultry may be contaminated with Serratia. Adeadly Serratia epizootic among chick embryoswas observed in a Japanese hatchery (Izawa etal., 1971). The hens carried S. marcescens in theirdigestive tract, but were themselves unaffected.Contamination of chicken carcasses with S. liq-uefaciens (Lahellec et al., 1975) and spoilage ofeggs by red-pigmented Serratia (Alford et al.,1950) have been reported. A disseminated sup-purative infection in a blue and gold macaw (Araararauna) affected with a chronic lymphoid atro-phy has also been recorded (Quesenberry andShort, 1983).

Serratia strains (mostly red-pigmented) areresponsible for 0.2–1.5% of cases of mastitis incows (Barnum et al., 1958; Roussel et al., 1969;Wilson, 1963). Raw milk, therefore, may occa-sionally contain Serratia spp., and S. liquefaciensand S. grimesii are common in dairy products(Grimes and Hennerty, 1931). Serratia strainshave been involved in septicemia in foals (Deomand Mortelmans, 1953), goats (Wijewanta andFernando, 1970), and pigs (Brisou and Cadeillan,1959); they have also been implicated in conjunc-tivitis of the horse (Carter, 1973) and abortion incows (Smith and Reynolds, 1970). The isolationof Serratia from the anal sac of the red fox Vulpesvulpes (Gosden and Ware, 1976) has beenreported.

Serratia strains have rarely been systematicallysearched for in the gut of animals. S. fonticola


was isolated in fecal samples from seven spar-rows and unidentified birds. The study involved90 wild European birds (Müller et al., 1986).

About 40% of trapped wild rodents andshrews carried Serratia strains in their gut with-out any visible sign of infection upon autopsy.The following animal species were found to carrySerratia spp.: 75/180, Apodemus sylvaticus; 7/23,Microtus arvalis; 3/11, Clethrionomys glareolus;1/2, Micromys minutus; (rodents); 7/9, Sorex; and1/3, Crocidura (shrews) (P. Giraud, F. Grimont,P.A.D. Grimont, unpublished observations). TheS. liquefaciens complex (S. liquefaciens, S. pro-teamaculans and S. grimesii) represented 73 to94% of all Serratia isolated from the gut of smallmammals and from the soil and plants aroundtraps (Table 1).

Serratia in HumansThe healthy human being does not often becomeinfected by Serratia, whereas the hospitalizedpatient is frequently colonized or infected. Atpresent, S. marcescens is the only known nosoco-mial species of Serratia (Table 1). S. liquefaciensand S. rubidaea are occasionally isolated fromclinical specimens, but their pathogenic role isnot established. The isolation of other Serratiaspecies is anecdotal (Farmer et al., 1985; Gill etal., 1981).

Clinically, Serratia infections do not differfrom infections by other opportunistic pathogens(von Graevenitz, 1977): respiratory tract infec-tion and colonization of intubated patients (e.g.,Cabrera, 1969; von Graevenitz, 1980); urinarytract infection and colonization of patients withindwelling catheters (e.g., Maki et al., 1973); sur-gical wound infection or superinfection (e.g.,Cabrera, 1969); and septicemia in patients withintravenous catheterization or complicating alocal infection (osteomyelitis, ocular or skininfections) (e.g., Altemeier et al., 1969). Menin-gitis, brain abscesses, and intraabdominal infec-tions are more exceptional.

The relationship between a Serratia strain anda patient may be in the form of an ephemeralassociation (in gut or throat, on hands or skin),a long-term colonization (in gut or urinary tractor on the skin), or a localized or generalizedinfection. The form of relationship might dependon the species or strain of Serratia, the entryroute (ingestion, injection, catheter), an ecologicadvantage (antibiotic treatment), or the patient’sphysiologic status. Patient factors have beenreviewed by von Graevenitz (1977). The localiza-tion of a hospital-acquired infection is oftendetermined by the kind of instrumentation orintervention done (i.e., the entry route). Thesame strain may cause a urinary infection in aurology ward, a bronchial colonization in an

intensive care unit, and a wound infection in asurgery unit. Five epidemiological situation mod-els can be described (adapted from Farmer et al.,1976):

Model 1: “Endogenous,” nonepidemic infec-tions. Sporadic cases of infection are observedthat are associated with different Serratia strains.The strains are often susceptible to several anti-biotics. The presence of a Serratia strain in fecesis not a sufficient proof of the endogenous originof the infection (Serratia are probably ingesteddaily with food). There is no prevention mecha-nism in this epidemiological model.

Model 2: Common source epidemics. A singlestrain (species, biotype, serotype) is found tocolonize or infect several patients. Any type ofSerratia can be involved, including pigmentedbiotypes of S. marcescens or environmental spe-cies (e.g., S. liquefaciens, S. grimesii, S. rubidaea).The strain is often susceptible to several anti-biotics. An investigation can reveal a commoninfection source such as a breathing machine(the nebulizer and tubings should be sampled),a batch of perfusion or irrigation fluid, or anantiseptic solution. Identification of the sourceusually allows efficient control of the epidemic.

Model 3: Patient-to-patient spread. The Serra-tia strain involved is typically a multiresistantmember of a nonpigmented biogroup of S.marcescens (biogroups A3, A4, A5/8, or TCT).No common source is found although severalsecondary sources are possible (sink or spongein patients’ rooms, high rate of fecal carriage).Disinfection of inanimate sources often has noeffect on the endemic state. In fact, the reservoiris usually the infected patient, and spreadamong patients occurs by transient carriage onthe hands of nursing or medical staff. Handlingof urinary catheters, wound drains, or trachealtubes of infected (or colonized) patients con-taminates the hands of personnel (Maki et al.,1973; Traub, 1972b). Hasty hand washing in anoverbusy ward or in the course of an emergency(for example, (in an intensive care unit) allowsthe transmission of the strain to uninfectedpatients. These patients are often immunocom-promised, treated preventively with broadspectrum antibiotics and subjected to diverseinstrumentation. The situation is typicallyendemic with epidemic peaks in periods of timewhen the ward is crowded. Transfer of infectedor colonized patients from one ward to anotheror from one hospital to another often results inthe spread of the outbreak to other wards orhospitals. Prevention of this epidemiologicalmodel is difficult (and in some countries, hope-less). Proper handwashing should be enforced.Some proposed solutions deal with ward/hospi-tal management (e.g., smaller wards, highernurse/patient ratio, separation of infected from


noninfected patients, and separation of theirrespective nurses).

Model 4: Colonization of the newborn intesti-nal tract. Typically, an investigation of a case ofSerratia infection leads to the discovery thatmost newborns are colonized by a red-pig-mented, drug-susceptible strain of S. marcescens.A common source can be identified (“sterile”water or oily antiseptic solution used to clean thebaby’s skin). Contamination may occur on thefirst day of life. Multiplication of the strain insoiled diapers may show a red discoloration (red-diaper syndrome). Control of this situation issometimes difficult due to the size of the reser-voir. Newly sterilized solutions are quickly con-taminated again. Enforcement of handwashingand frequent sterilization (daily or more) ofincriminated solutions may be helpful.

Model 5: Pseudoepidemics. A drug-sensitivestrain of any Serratia species (e.g., S. liquefaciens)is unreproducibly isolated from several patientswho show no sign of infection. Investigation ofthe plastic material used to “sterilely” collectblood occasionally allows the isolation of theenvironmental strain that contaminated thesystem (plastic tubing, EDTA, or citratesolution).

The above situations are sometimes mixed orless clear. References to reports fitting with theabove models can be found in Farmer et al., 1976;Schaberg et al., 1976; von Graevenitz, 1977, 1980;and Daschner, 1980.

Properties Relevant to Pathogenicity in HumansSerratia marcescens is generally an opportunisticpathogen causing infections in immunocompro-mised patients. Among the possible pathogenic-ity factors found in Serratia strains are theformation of fimbriae, the production of potentsiderophores, the presence of cell wall antigens,the ability to resist to the bactericidal action ofserum, and the production of proteases.

In practice, each strain of the genus Serratiaproduces one to three different kinds of fimbrialhemagglutinin (HA) (Old et al., 1983). Fivetypes of fimbriae have been observed inserratiae:

Type 1 fimbriae: thick, channelled fimbriae ofexternal diameter 8 nm, associated with a man-nose-sensitive hemagglutinin (MS-HA) reactingstrongly with untanned fowl or guinea pig eryth-rocytes. The production of MS-HA is increasedby serial, static broth cultures in air at either 20,30, or 37°C. Production of MS-HA was found tobe correlated with the ability of S. marcescenscells to attach to human buccal epithelial cells(Ismail and Som, 1982) or to the human urinarybladder surface (Yamamoto et al., 1985). This

type of HA was found to be produced by all (Oldet al., 1983) or almost all (Franczek et al., 1986)S. marcescens strains, whether environmental orclinical, and in some strains of other Serratia spe-cies, except S. plymuthica and S. fonticola.

Type 3 fimbriae: thin, non-channelled fimbriaeof external diameter 4–5 nm associated with amannose-resistant hemagglutinin reacting withtannic acid-treated, but not fresh, oxen erythro-cytes (MR/K-HA) (Old et al., 1983). MR/K-HAwas found to be produced by almost all strainsof all Serratia species studied by Old et al. 1983.However, Franczek et al. 1986 found MR/K-HAwas more frequently produced by clinical thanby environmental strains of S. marcescens. TheMR/K-HA of all Serratia species, except S.rubidaea, were immunologically related. MR/K-HA from S. rubidaea was immunologicallyrelated to a Klebsiella MR/K-HA (Old et al.,1983).

Thin fimbriae associated with a mannose-resis-tant hemagglutinin reacting with fowl, guineapig, and horse erythrocytes (type FGH MR/P-HA). This HA was produced by strains from allspecies except S. plymuthica, S. odorifera, and S.fonticola (Old et al., 1983). The correspondingfimbriae are immunologically related in the dif-ferent species.

Thin fimbriae associated with a mannose-resistant hemagglutinin reacting with fowl eryth-rocytes only (type F MR/P-HA). This HA wasproduced by some S. rubidaea strains and wasimmunologically unrelated to other hemaggluti-nins (Old et al., 1983).

Thick, channelled fimbriae of external diame-ter 9–10 nm: associated with a mannose-resistanthemagglutinin reacting with fowl erythrocytesonly (type F MR/P-HA). This HA was producedby some S. fonticola strains (Old et al., 1983).

Nearly all clinical or environmental S. marce-scens strains produce potent siderophore(s)capable of scavenging iron from ethylenedi-amine di-O-hydroxyphenylacetic acid, a chelatorwith an association constant for ferric iron of1033.9 (Franczek et al., 1986). Serratia strains (S.marcescens and S. liquefaciens were tested) gen-erally produce enterobactin (Reissbrodt andRabsch, 1988) but only rarely produce aerobac-tin (Martinez et al., 1987). A novel iron (III)transport system named SFU, was evidenced ina S. marcescens strain. In this system, no sidero-phore production is involved (Zimmermann etal., 1989).

At least two patterns of hemolysis were foundto be produced by S. marcescens on horse blood:a clear-cut narrow zone of hemolysis underthe colony, evoking the action of a cell-boundhemolysin, and a fuzzier, diffusing zone ofhemolysis evoking the production of a solublehemolysin (Grimont and Grimont, 1978b). The


latter hemolysis pattern was produced only by S.marcescens biogroup A4.

The cell-bound hemolysin has been studiedand direct contact between erythrocytes and S.marcescens cells has been demonstrated (Braunet al., 1985). The lysis of erythrocytes requiresactively metabolizing bacteria and does not needcalcium ions. A 7.3-kilobase-pair chromosomalfragment encoding the hemolytic activity hasbeen cloned in Escherichia coli and sequenced(Poole et al., 1988). Two open reading framesdesignated shlA and shlB have been observed.The hemolysin protein (molecular weight165,056) is encoded by shlA, and the product ofshlB somehow activates shlA. Protein shlA inte-grates into the erythrocyte membrane and causesosmotic lysis through channel formation (Schie-bel and Braun, 1989). When an E. coli straincarrying the Serratia hemolysin gene was com-pared to the isogenic strain without such a genein an experimental rat model (107 CFU/mlinjected via urethra into the bladder), thestrain carrying the hemolysin gene colonized theurinary tract more and led to a stronger inflam-matory response compared to the hemolysin-negative strain (Marre et al., 1989).

A capsulated Serratia strain injected in theperitoneal cavity of the mouse multiplied andkilled the mouse whereas an uncapsulated vari-ant was avirulent (Ohshima et al., 1984).

Traub (1983) has shown that antibodiesdirected against O antigens of challenge strainsafford passive protection in mice. The predomi-nance of the O6 and O14 serotypes of S. marce-scens in infections, the surface localization ofLPS and its masking of other antigens suggeststhat LPS is a prime candidate for exploitation asa protective antigen in a vaccine against S.marcescens (Jessop, 1985).

Traub and Fukushima (1979a) classified S.marcescens strains into three categories withrespect to their serum susceptibility. In solutionswith 80% (vol/vol) of fresh serum, 88% of strainswere found to be “delayed serum-sensitive” i.e.,they were killed after a few hours exposure; 6%of strains were “promptly serum-sensitive” i.e.,they were killed within a matter of minutes; and6% of strains resisted an overnight exposure toserum. The delayed serum-sensitive strains wereshown to be killed via the activation of the alter-native pathway of the human complement sys-tem since killing was unaffected by inhibition ofthe classical pathway (Traub and Kleber, 1976).Conversely, promptly serum-sensitive strainswere killed in a delayed fashion when the classi-cal pathway was inhibited (Traub and Kleber,1976). Depletion of fresh human serum of C3with hydrazine hydrate completely abolished thebacterial activity against both promptly serum-sensitive and delayed serum-sensitive strains

(Traub and Fukushima, 1979a). Complement isof prime importance with respect to efficientopsonization and subsequent phagocytic killingby human peripheral blood granulocytes (Traub,1982). In an intraperitoneal mouse model, treat-ment of animals with cyclophosphamide (gener-ating a leukopenia) or zymosan (depletion ofcomplement) did not increase susceptibility to S.marcescens. A combination of both treatmentswas necessary to significantly raise susceptibilityto S. marcescens (Traub et al., 1983).

Crude culture filtrates have been shown toproduce dermal hemorrhage after intradermalinoculation (Liu, 1961), corneal damage (Kregerand Griffin, 1975), and endophthalmitis afterintravitreal injection (Salceda et al., 1973). Theseare due to one or more extracellular proteasesproduced by S. marcescens. Up to four proteaseswere purified (Lyerly and Kreger, 1979; Matsu-moto et al., 1984)—two metalloproteases of 56and 60 KDa and two thiol proteases of 73 KDa.These proteases play a prominent role inthe pathogenesis of experimental pneumonia(Lyerly and Kreger, 1983) and keratitis (Lyerlyet al., 1981). Most studies focused on the 56-KDaprotease. Protease(s) were shown to cause: 1)liquefactive necrosis of the cornea (Kamata etal., 1985); 2) inactivation of five major proteinaseinhibitors including "1-proteinase inhibitor(Virca et al., 1982), "2-macroglobulin, Cl inhibi-tor, "2-antiplasmin, and antithrombin III (Mollaet al., 1989), which participate in regulating var-ious cascade mechanisms (fibrinolysis and clot-ting cascade, complement system, inflammatoryresponse); 3) degradation of immunoglobulins Gand A, fibronectin, and other serum proteins(Molla et al., 1986; Traub and Bauer, 1985); 4) apotent toxic effect of fibroblasts that was medi-ated by internalization of a proteinase-"2-macro-globulin complex via the "2-macroglobulinreceptor, followed by regeneration of the pro-teinase activity in the cell (Maeda et al., 1987);5) activation of the Hageman factor-kallikreinsystem generating kinin, which leads toenhanced vascular permeability (Matsumoto etal., 1984); and 6) cleavage of human C3 and C5to yield leukotactic fragments (Ward et al.,1973).

Although resistance plasmids probably playno role in the virulence of S. marcescens in exper-imental models (Traub et al., 1983), multipledrug resistance may affect the course and prog-nosis of infections.

Non-pigmented strains of S. marcescens aregenerally more resistant to antibiotics thanpigmented strains because they often harborresistance plasmids. Environmental strains of S.marcescens or strains isolated before the antibi-otic era are resistant to colistin, cephalothin,ampicillin (low level of resistance), tetracycline,


and nitrofurantoin and are usually susceptible tocarbenicillin, third-generation cephalosporins,chloramphenicol, streptomycin, kanamycin, gen-tamicin, tobramycin, amikacin, trimethoprim/sulfamethoxazole, fosfomycin, nalidixic acid, andother quinolones (P.A.D. Grimont, unpublishedobservations). Clinical strains harboring plas-mids often show additional resistance to variousantibiotics (Farrar, 1980; Hedges, 1980), with asingle plasmid being able to confer resistance toone to eleven antibiotics (Hedges, 1980; Olexy etal., 1982). Mutants resistant to nalidixic acid areoften encountered in urology wards, and thesemutants are often resistant to other quinolones.The antibiotics most often active against nosoco-mial strains of S. marcescens are amikacin, mox-alactam, and cefotaxime. However, some strainsmay produce enzymes inactivating amikacin(Farrar, 1980) or may overproduce a class C #-lactamase (low-level resistance to cefotaximeand some other third-generation cephalospor-ins), which may combine with decreasedpermeability to #-lactams (high resistance tocefotaxime and some other third-generationcephalosporins) (Hechler et al., 1989).

Compounds Produced by Serratia

PigmentsProdigiosin, a nondiffusible red pigment, is a sec-ondary metabolite formed by the enzymaticcondensation of 2-methyl-3-amylpyrrole (MAP)and 4-methoxy-2,2$-bipyrrole-5-carboxyaldehyde(MBC), leading to a tripyrrole derivative, 2-methyl-3-amyl-6-methoxyprodigiosene (Wil-liams and Qadri, 1980). Little is known about thebiosynthetic pathways of MAP or MBC, exceptthat a proline molecule is incorporated intactinto one of the pyrrole groups of MBC (Williamsand Qadri, 1980). Pigment synthesis require air,probably molecular oxygen. In the genus Serra-tia, prodigiosin is only produced by strains of S.marcescens, S. plymuthica, and S. rubidaea. In S.marcescens, prodigiosin is produced by bio-groups A1 and A2/6 and never by biogroups A3,A4, A5/8, or TCT (Grimont, P. A D., 1977). Non-pigmented strains of biogroups A1 or A2/6 areoften blocked in the synthesis of either MAP orMBC (Grimont, P. A. D., 1977; Williams andQadri, 1980). Strains in the nonpigmented bio-groups are likely to lack the condensing enzyme(Ding and Williams, 1983). Some genes encodingprodigiosin biosynthesis were cloned by Dauen-hauer et al. 1984 and expressed in Escherichiacoli. The clones obtained had acquired the abilityto condense MAP with MBC or to produce MAPin addition to the condensing enzyme.

Strains belonging to the ubiquitous biotype A4can synthesize a pink diffusible pigment called

pyrimine (Grimont and Grimont, 1984; Williamsand Qadri, 1980). Pyrimine, or ferrorosamine A,is a ferrous complex of L-2(2-pyridyl)-D$-pyrro-line-5-carboxylic acid, a secondary metabolitealso known to be produced by Erwinia rhapon-tici (Feistner et al., 1983).

A yellow diffusible pigment, 2-hydroxy-5-carboxymethylmuconic acid semialdehyde, isproduced from the meta cleavage of 3,4-dihydroxyphenylacetic acid (3,4-DHP) by theenzyme 3,4-DHP 2,3-dioxygenase (Trias et al.,1988). This enzyme is induced by tyrosine in allS. marcescens strains. However, presently only S.marcescens strains of biotype A8a, which havelost the ability to grow on aromatic compounds,can produce the yellow pigment. An uncoloredmuconic acid, #-cis-cis-carboxymuconic acid,can be produced by the ortho cleavage of 3,4-dihydroxybenzoate.

BiosurfactantsBoth pigmented and some nonpigmented strainsof Serratia marcescens produce a biosurfactantthat can act as a wetting agent. This wetting agentis produced in large amounts at 30°C but not at37°C during the stationary phase of growth.Three aminolipids, designated W1 to W3 andhaving the wetting activity, were separated bythin-layer chromatography (Matsuyama et al.,1986). W1 was identified as serratamolide, acyclodepsipeptide earlier discovered by Wasser-man et al. 1961.

Fatty AcidsThe major fatty acids components in whole-cellmethanolysates of Serratia species are 3-hydroxy-tetradecanoic, n-hexadecanoic, hexadecanoic,and octadecanoic (not separated from octadeca-dienoic) acids. These contribute 50–80% of thecomponent fatty acids in each strain. Significantquantities are also contributed by n-tetradecanoic acid (Bergan et al., 1983).

FlavorsAlkyl-methoxypyrazines are responsible for thepotatolike odor produced by all strains of S.odorifera and S. ficaria and some strains of S.rubidaea. The major alkyl-methoxypyrazine pro-duced by these Serratia strains is 3-isopropyl-2-methoxy-5-methylpyrazine. Minor compoundsinclude 3-sec-butyl-2-methoxy-5(6)-methylpyra-zine and 3-isobutyl-2-methoxy-6-methylpyrazine(Gallois and Grimont, 1985). Production ofthese three molecules by other bacterial specieshas never been observed. The food industryuses pyrazines to improve the flavor of someproducts.


EnzymesThe most probable potential use of chitinase isin the treatment of the chitin-containing wastesproduced by the seafood packing industry, bio-control agents against plant pathogenic fungi,production of adhesives and wound dressings,heavy metal recovery from waters, delayed-release agrochemicals or drugs, and dialysismembranes (Joshi et al., 1989). S. marcescensproduces five proteins (molecular masses of 57,52, 48, 36, and 21 kDa) with chitinolytic activity(Fuchs et al., 1986) whereas S. liquefaciens pro-duces three proteins (molecular masses of 56, 51,and 36 kDa) with such activity (Joshi et al., 1989).The sequences of two S. marcescens chitinasegenes, chiA encoding the 57 kD enzyme andchiB encoding the 52 kD enzyme, have beenpublished (Jones et al., 1986; Harpster andDunsmuir, 1989).

BiotechnologySerratia strains have been used in whole-cellbioconversion processes. Resting-cell suspen-sions of S. marcescens are able to convertvanillin to vanillic acid. At high substrate con-centration (0.3%), 75% vanillin is converted(Perestello et al., 1989). 2,5-Diketogluconic acid,a key intermediate in the synthesis of ascorbicacid, can be produced from glucose at 20°C byS. marcescens (without requirement for a cofac-tor) and by S. liquefaciens and S. grimesii whensupplied with the cofactor, pyrroloquinolinequinone (Bouvet et al., 1989).

Some genetic tools are available to constructS. marcescens strains of biotechnological interest.These tools include general transduction (Matsu-moto et al., 1973), transformation (Reid et al.,1982), and use of nuclease-deficient, antibiotic-sensitive, and restrictionless mutants (referenceis not an exact match Takagi and Kisumi, 1985).Recombinant strains obtained were histidase-less regulatory mutants producing L-histidine(Kisumi et al., 1977), L-arginine-producingmutants (Kisumi et al., 1978), an isoleucine-pro-ducing strain (Komatsubara et al., 1980), and athreonine-hyperproducing strain (Komatsubaraet al., 1983).

Isolation

Underlying PrinciplesPigmented species and biotypes of Serratia oftenexhibit pink or red colonies on nutrient agar.Use of low-phosphate agar without glucose,such as peptone-glycerol agar (Difco peptone,

5 g; glycerol, 10 ml; Difco agar, 20 g; distilledwater, 1 liter), is best in order to demonstratepigmentation (Williams and Hearn, 1967).Although the genus Serratia includes all the red-pigmented enterobacteria, identification of apink or red colony should be confirmed by bio-chemical tests because a few nonenterobacteriamay also produce prodigiosin (see “Introduc-tion”). On nutrient agar, nonpigmented speciesor biotypes of Serratia give opaque-whitish,mucoid, or transparent smooth colonies. Noneof these traits is specific for the isolation ofSerratia from a mixture of enteric bacteria.Colonies of S. odorifera and S. ficaria andoccasionally S. rubidaea give off a specific,potato-like odor. Detection of this odor maysuggest the presence of at least one colony ofthat species on the plate.

The salt tolerance and relatively low minimalgrowth temperature of all Serratia species mayhelp devise a means of enriching them. A nutri-ent broth containing 4% NaCl, incubated at 15–20°C, will allow multiplication of Serratia but notof Enterobacter cloacae or Aeromonas hydro-phila (P. A. D. Grimont, unpublished observa-tions). This procedure has been used by us, butno systematic study of its effectiveness has beenconducted.

Strains of the genus Serratia do not normallyrequire addition of growth factors to a minimalmedium. Thus, an enrichment medium for Serra-tia can be a minimal medium with a carbonsource that is commonly used by Serratia spp. butnot by non-Serratia spp.

Production of extracellular gelatinase, lecithi-nase, and DNase are essential characteristics ofthe genus Serratia (Ewing, 1986; Grimont et al.,1977b), S. fonticola excepted. Agar media thatshow the presence of one or a combination ofextracellular enzymes can be used to differenti-ate Serratia colonies.

Finally, Serratia spp. are resistant to severalcompounds, including colistimethate, cephal-othin (Greenup and Blazevic, 1971), and thal-lium salts (Starr et al., 1976), and one of thesecompounds can be used to make enrichment ordifferential media selective for Serratia. How-ever, some S. plymuthica or S. rubidaea strainsmay be more susceptible to colistimethate, ceph-alothin, or ampicillin than is S. marcescens.

No medium has yet been devised to selectivelyisolate one particular Serratia species (other thanS. marcescens), although specific carbon sourcesor conditions are listed in taxonomic papers(Grimont et al., 1977b). Occasionally, a selectivemedium may be devised according to a specificpattern of antibiotic resistance displayed by anosocomial strain of S. marcescens (Denis andBlanchard, 1975), but such media cannot be ofgeneral utility.


Selective Media Based on DNase Production and Antibiotic ResistanceFarmer et al. 1973 proposed the following deoxy-ribonuclease-toluidine blue-cephalothin (DTC)agar.

DTC Agar for Isolating Serratia (Farmer et al., 1973)Deoxyribonuclease test agar (BBL) 21 gAgar (Difco) 2.5 gToluidine blue O 0.05 gDistilled water 500 ml

Mix on a mechanical stirrer until the dye dissolves, auto-clave with a Teflon stirring bar at 121°C for 15 min, coolto 50°C, add 5 ml (500 mg) of cephalothin (Keflininjectible, Eli Lilly), stir and dispense in sterile petri dishes.

Typically, Serratia spp. grow on this blue DTCmedium producing a red halo extending severalmillimeters around the colonies. “False nega-tives” (no growth or no halo) are very rareamong S. marcescens isolates, but several strainsof S. rubidaea and S. plymuthica failed to give theproper reaction (Starr et al., 1976).

A medium with DNase test agar, toluidineblue, cephalothin (30 µg/ml and colistimethate(30µ/ml) has been proposed by Cate (1972); anda medium with DNase test agar, toluidine blue,egg yolk, and cephalothin (100µ/ml) has beenproposed by Goldin et al. 1969. The latter com-bines DNase and lecithinase detection.

Berkowitz and Lee (1973) devised the follow-ing medium for the isolation of S. marcescens:

DNase Medium for Isolating Serratia marcescens(Berkowitz and Lee, 1973)

Deoxyribonuclease test agar with methyl green (Difco) 42 g

L-Arabinose 10 gPhenol red 0.05 gMethyl green, 1% 4 mlDistilled water up to 1 liter

After autoclaving, add ampicillin (5 µg/ml), colistime-thate (5 µg/ml), cephalothin (10 µg/ml), and amphotericinB (2.5 µg/ml).

Serratia colonies hydrolyze DNA, and thegreen component of the medium’s dark colordisappears around the colonies. S. marcescens (orS. entomophila), unable to ferment L-arabinose,will give colonies surrounded by a red halo,whereas other Serratia species will give a yellowhalo. The other Serratia species, however, may beinhibited by the antibiotic mixture. Wright et al.1976 used this medium to isolate and enumerateS. marcescens from vegetable salads.

Selective Media Based on Carbon-Source UtilizationA minimal medium with meso-erythritol asthe sole carbon source was devised by Slotnickand Dougherty (1972). A modification of this

medium that included the antiseptic “Irgasan”(4$,2$,4$-trichloro-2-hydroxydiphenylether), hasbeen proposed (Lynch and Kenealy, 1976).Unfortunately, Serratia liquefaciens, S. plymuth-ica, S. entomophila, S. proteamaculans biotypesClc and RB, S. odorifera biotype 1, and the noso-comial biotypes A5, A8a, A8b, A8c and TCT ofS. marcescens cannot grow with erythritol as solecarbon source (Grimont et al., 1977b; Grimontet al., 1978a; Starr et al., 1976). Therefore, thesemedia should not be used in epidemiological orecological surveys unless the study is knowinglylimited to erythritol-positive serratiae.

A minimal medium with meso-inositol as solecarbon source has been proposed for the selec-tive isolation of Klebsiella pneumoniae and Ser-ratia spp. (Legakis et al., 1976). However, only afew genera and species have been studied bythese authors, and it has been shown (Grimontet al., 1977b) that Enterobacter aerogenes, E. clo-acae, Erwinia herbicola, and pectinolytic Erwiniaspp. can also grow on a minimal medium withinositol as sole carbon source.

A caprylate-thallous (CT) agar medium hasbeen devised for the selective isolation of all Ser-ratia species and biotypes. This CT agar, derivedfrom M70 minimal medium (Véron, 1975), ismade up as follows (Starr et al., 1976):CT Agar for Selective Isolation of Serratia(Starr et al., 1976)

First, a trace element solution (Véron, 1975) is prepared:

H3PO4 1.96 gFeSO4·7H2O 0.0556 gZnSO4·4H2O 0.0287 gMnSO4·4H2O 0.0223 gCuSO4·5H2O 0.025 gCo(NO3)2·6H2O 0.003 gH3BO3 0.0062 gDistilled water 1 liter

Store at 4°C (keeps well for at least 1 year). Thenthe following solutions (solution A and solution B) areprepared:

Solution A:

CaCl2·2H2O 0.0147 gMgSO4·7H2O 0.123 gKH2PO4 0.680 gK2HPO4 2.610 gTrace element solution (see above) 10 mlCaprylic acid 1.1 mlYeast extract (Difco)

(5% wt/vol solution) 2 mlThallous sulfate 0.25 gDistilled water up to 500 ml

The pH is adjusted to 7.2 with NaOH. Autoclave at 110°C(or 120°) for 20 min.

Solution B:

NaCl 7 g(NH4)2SO4 1 gAgar (Difco) 15 gDistilled water 500 g


The pH is adjusted to 7.2. Autoclave at the same time assolution A. After autoclaving, solutions A and B aremixed aseptically, and the resulting medium is poured inthick layers into sterile, plastic petri dishes (25–30 ml forpetri dishes 9–10 cm in diameter). The medium is nolonger effective if it is remelted, but plates of CT agarkeep well for several weeks at 4°C if contamination anddesiccation are prevented.

CT agar is very useful for the isolation of Ser-ratia spp. (except S. fonticola) from feces, spu-tum, and any other polymicrobial clinical sample.Serratia colonies are apparent within 3 days, andfurther incubation allows colonies grow (2–5mm). Occasionally, Providencia spp., Acineto-bacter spp., or Pseudomonas spp. may developcolonies on this medium. Other bacteria giveonly pinpoint colonies. When samples are fromthe natural environment or food, the results areless clear, especially if the sample contains nutri-ents. Several colonies must then be checked forDNase. Preenrichment in nutrient broth with4% NaCl, incubated overnight at 20°C, can beused; about 0.1 ml of the enrichment culture isstreaked onto CT agar. Broth that has supportedovernight bacterial growth usually does notcontain enough nutrients to adversely affectthe selective isolation of Serratia spp. (P. A. D.Grimont, unpublished observations).

Identification

The usual methods for the identification ofserratiae and other Enterobacteriaceae can befound in Ewing (1986). However, following thepioneering example of Stanier et al. (1966) withthe pseudomonads, developments in enterobac-terial taxonomy (Bouvet et al., 1985; Grimontand Ageron, 1989; Grimont et al., 1977b; Gri-mont et al., 1988) have demonstrated the useful-ness of carbon source utilization tests. Also, itshould be noted that utilization tests are prefer-able to fermentation tests since strains able toutilize a polyalcohol sometimes fail to produceenough acid products to give a positive reactionin fermentation tests. Although all species can beidentified with carbon source utilization tests,some conventional tests can also be used:tetrathionate-reduction and #-xylosidase testsare useful in clinical identification (Le Minor etal., 1970; Brisou et al., 1972). The methodologyof carbon source utilization tests and other pro-cedures that are not in general use, but are essen-tial to Serratia identification, are detailed herein.

Conditions of IncubationBest results are obtained when Serratia culturesare incubated at 30°C. At 37°C, pigmentationoften fails to appear, and the Voges-Proskauer

test is often negative. S. plymuthica may even failto grow at 37°C. Otherwise, there is little differ-ence between the results of tests held at 30°Cand those held at 37°C as far as S. marcescens isconcerned.

Carbon Source Utilization TestsTwo methods can be used for carbon utilizationtests: a minimal agar medium or API strips (APISystem, La Balme-les-Grottes, France).

The minimal agar medium employs the sameM70 medium (Véron, 1975) used in the capry-late-thallous selective agar (see the recipe), butomits the yeast extract, thallous sulfate, andcaprylate from solution A. Neutralized carbonsources (the purest brand available) are addedto solution A to give a 0.1% final concentration(weight of the anion/vol) in the completemedium (0.2% in the case of carbohydrates).Sterilization of the carbon source in aqueoussolution is achieved by filtration throught mem-brane filters or by heating at 80°C for 20 min.Plates are examined for growth after 4 days(early reading) and 14 days (late reading).

The API strip method uses API 50 CH (con-taining 49 carbohydrates), API 50 AO (49organic acids), and API 50 AA (49 aminoacids)galleries and a minimal medium containinggrowth factors provided by the manufacturer.These were described by Gavini et al. (1980).Since about one-third of the carbon sources areuseless in the study of the Enterobacteriaceae,we use special galleries containing 99 selectedcarbon sources (including new compounds notincluded in the above-mentioned galleries). Theinoculated galleries are examined for growthafter 2 days (early reading) and 4 days (late read-ing). These galleries have been used for thedescription of several bacterial groups (Bouvetet al., 1985; Grimont et al., 1988; Grimont andAgeron, 1989).

In our experience, carbon source utilizationtests (CSUT) gave reproducible and clear-cutresults. Utilization of carbohydrates and polyal-cohols often give more useful results than fer-mentation tests (FT). More than 95% of CSUTand FT using D-xylose, L-arabinose, trehalose,sorbitol, sucrose, and rhamnose gave the sameresults (P. A. D. Grimont, 1977). However,results with CSUT and FT are not parallel whenadonitol, inositol, cellobiose, and lactose arestudied. Whereas almost all Serratia strains cangrow on inositol, acid from inositol was producedby only 76% of S. marcescens, 36% of S.rubidaea, 92% of S. liquefaciens-S. proteamacu-lans and 35% of S. plymuthica strains within 2days. CSUT with adonitol is a very good test forseparating S. marcescens from S. liquefaciens(99% vs. 0% positive strains, respectively), but


FT with adonitol is a poor test for separatingthese two species (32% vs. 0% positive strains in2 days, respectively).

Glucose Oxidation TestThe Lysenko test (Lysenko, 1961) was exten-sively modified by Bouvet et al. 1989. A 1-literportion of glucose oxidation medium is com-posed of basal medium containing 8 g of nutrientbroth (Difco), 0.02 g of bromcresol purple, and0.62 g of MgSO4·7H2O (2.5 mM). To 9.2 ml ofautoclaved (121°C, 15 min) basal medium isadded 0.4 ml of a filter-sterilized 1 M D-glucosesolution (final concentration, 40 mM). Thismedium is supplemented with 0.4 ml of a fresh,sterile 25 mM iodoacetate solution (final concen-tration, 1 mM). The glucose oxidation medium isdistributed (0.5 ml portions) into glass tubes (11by 75 mm), which are plugged with sterile cottonwool. Bacteria grown overnight at 20 or 30°Con tryptocasein soy agar (Diagnostics Pasteur,Marnes-la-Coquette, France) supplemented with0.2% (wt/vol) D-glucose are collected with aplatinum loop, suspended in a sterile 2.5 mMMgSO4 solution and adjusted to an absorbance(at 600 nm) of about 4. Glucose oxidationmedium is then inoculated with 50µ1 of bacteriasuspension and vigorously shaken at 270 strokesper minute overnight at 20 or 30°C. The test ispositive if a yellow color develop (acid produc-tion). Otherwise, the medium remains purple.When negative, the glucose oxidation test isrepeated in the presence of 10µM pyrroloquino-line quinone (PQQ). The control strains used arethe type strains of S. marcescens (positive with-out requirement for PQQ) and S. liquefaciens(positive only when PQQ is supplied) (Bouvetet al., 1989).

Gluconate and 2-Ketogluconate Dehydrogenase TestsThe gluconate dehydrogenase test is performedas follows (Bouvet et al., 1989): Bacteria aregrown overnight at 20°C on tryptocasein soyagar supplemented with 0.2% D-gluconate, thencollected with a platinum loop and suspended insterile distilled water to an absorbance (at 600nm) of about 4. The reaction medium contains0.2 M acetate buffer (pH5), 1% (wt/vol) TritonX-100, 2.5 mM MgSO4, 75 mM gluconate, and(added immediately before use) 1 mM iodoace-tate. A control medium contains the same ingre-dients except gluconate. The reaction mediumand the control medium are dispensed (100µlportions) into 96-well microtiter plates (Dyna-tech AG, Denkendorf, Germany). Bacterial sus-pensions (10µl) are added to the reaction and the

control media, and the microtiter plates are incu-bated at 20°C for 20 min. Then, 10-µl portions ofa 0.1 M potassium ferricyanide solution (kept inthe dark at room temperature for no longer than1 week) are added to the wells, and the plates aregently shaken and incubated at 20°C for 40 min.Then 50 µ1 of a reagent containing Fe2(SO4)3, 0.6g; sodium dodecyl sulfate, 0.36 g; 85% phospho-ric acid, 11.4 ml; distilled water to 100 ml is addedto each well. The plates are examined for thedevelopment of a green- to-blue color (due toPrussian blue) within 15 min at room tempera-ture. The color in the uninoculated controlmedium remains yellow. The suggested controlstrains used are Escherichia coli K-12 (negative)and the type strain of Serratia marcescens(positive).

The 2-ketogluconate dehydrogenase test isthe same as above, except that the control andreaction media are adjusted to pH 4.0, and 2-ketogluconate is used in place of gluconate inthe reaction medium. Suggested control strainsused are Escherichia coli K-12 (negative) and thetype strain of Serratia marcescens (positive at20°C, not at 30°C) (Bouvet et al., 1989).

Tetrathionate Reductase TestThe tetrathionate reductase test was found to bevery useful for the identification of Serratia spp.The liquid medium of Le Minor et al. (1970) isdescribed below:

Tetrathionate Reductase Test for Identifying Serratiaspp. (Le Minor et al., 1970)

K2S4O6 5 gBromthymol blue (0.2%

aqueous solution) 25 mlPeptone water (peptone [Difco], 10 g;

NaCl, 5 g; distilled water, 1 liter) Up to 1 liter

Adjust pH to 7.4. Filter-sterilize and dispense 1 ml into12-x-120-mm tubes. Within 1 or 2 days, reduction oftetrathionate will acidify the medium, indicated by a yel-low color. Negative tubes remain green or turn blue.

Voges-Proskauer TestResults of the Voges-Proskauer test are variablewith some Serratia spp., probably because theend products of the butanediol pathway aremetabolized (Grimont et al., 1977b). Manystrains of S. plymuthica and S. odorifera give neg-ative results with O’Meara’s method (O’Meara,1931) and positive results with Richard’s proce-dure (Richard, 1972). In Richard’s procedure,0.5-ml aliquots of Clark and Lubs medium(BBL) are inoculated in 16-mm-wide test tubes.After overnight incubation at 30°C, 0.5 ml of "-naphthol (6 g in 100 ml absolute ethyl alcohol)and 0.5 ml of 4 M NaOH are added. Positivetubes will turn red.


Gas ProductionGas production differentiates Serratia marce-scens from S. liquefaciens—S. proteamaculansbetter when glucose agar is used than when aliquid medium with a Durham inverted tube isused (Grimont et al., 1977b). Glucose agar con-sists of nutrient agar (meat extract [Liebig], 3 g;yeast extract [Difco], 10 g; agar [BBL], 15 g; dis-tilled water to 1 liter; pH 7.4) supplemented with1% (wt/vol) glucose. Dispense into tubes (160 !10 mm). Autoclave 20 min at 120°C, cool at 50°C,and inoculate before the agar sets. Tubes areexamined for bubbles of gas for up to 3 days.

#-XylosidaseThe method of Brisou et al. (1972) for demon-strating #-xylosidase can conveniently beadapted for use with sterile microculture plates.Aqueous (1% wt/vol) p-nitrophenyl-#-xylosideis dispensed into the wells in 0.05-ml amounts,followed by 0.05 ml of a fresh bacterial suspen-sion in 0.25 M phosphate buffer, pH 7. Plates areexamined for a yellow color after 24 h.

Identification of Serratia at the Genus LevelMembers of the genus Serratia share the charac-teristics defining the family Enterobacteriaceae.Only occasionally can a nitrate-negative strainbe isolated. The properties that best define thegenus Serratia are listed in Table 3. Althoughlipase activity on tributyrin or corn oil is listed,S. odorifera strains are only weakly lipolytic. Aweak urease activity, lack of motility, or presenceof a capsule are occasionally observed. Serratiaeare clearly differentiated from Klebsiella spp.,Enterobacter aerogenes, and E. cloacae by pro-duction of gelatinase, lipase, DNase, and bygrowth on caprate or caprylate as sole carbonsource. Some soft-rot Erwinia spp. produceextracellular proteinase and DNase, but thesestrains are pectinolytic and lack glucose and glu-conate dehydrogenases (Grimont et al., 1977b).

Identification of Serratia SpeciesThe characteristics best allowing identification ofeach Serratia species are given in Tables 4 and 5.Carbon source utilization tests are invaluable forunambiguous identification. Indole productionby S. odorifera is not reproducibly observedwhen peptone water is used. A defined mediumcontaining trytophane (e.g., “urée-indole”medium, Diagnostics Pasteur, Marnes-la-Coquette, France; or the indole test in API 20Estrips) gives consistently positive results with thisspecies. With the use of classical tests (Ewing,1986), S. plymuthica may be confused either with

S. proteamaculans, S. liquefaciens, S. rubidaea,or Enterobacter agglomerans. In the past, S.plymuthica had been identified as “atypical S.liquefaciens” (negative for lysine and ornithinedecarboxylase) or atypical S. rubidaea (sorbitolpositive). The species S. odorifera can also berecognized in a number of “rhamnose-positive S.liquefaciens” strains. Carbon source utilization

Table 3. Characteristics defining the genus Serratia.a

aTaken from Grimont et al., 1977b, and unpublishedobservations.bMore than 90% of the isolates of each species give the samereaction.cONPG, ortho-nitrophenyl-galactoside.In the presence of pyrroloquinoline quinone.

Positive characteristics of all species of the genus Serratiab

Motile rodsGrowth at 20°C in 1 dayGrowth at pH 9Growth with 4% NaClGrowth on minimal medium without addition of growth

factorAcid from maltoseAcid from mannitolAcid from salicinAcid from trehaloseONPGc hydrolyzedGlucose oxidized to gluconated

Gluconate oxidized to 2-ketogluconateCarbon source utilization tests: N-acetylglucosamine, cis-

aconitate, 4-aminobutyrate, citrate, fructose, galactose, galacturonate, gluconate, glucose, glucuronate, glycerol, m-inositol, 2-ketogluconate, L-malate, maltose, mannitol, mannose, putrescine, ribose, trehalose

Additional positive characteristics of Serratia species other than S. fonticolab

Voges-Proskauer test (Richard)Lipase (tributyrin, corn oil)Growth on caprylate-thallous agarDNaseProteinase(s)

Negative characteristics of all species of the genus Serratiab

Urease (Ferguson)H2S from thiosulfatePhenylalanine/tryptophane deaminasePolygalacturonidaseAmylase (4-day reading)#-GlucuronidaseCarbon source utilization tests: 5-aminovalerate, m-

coumarate, ethanolamine, glutarate, histamine, L-sorbose, tryptamine

Sodium ion requirementAnaerobic growth with KClO3

Additional negative characteristics of Serratia species other than S. fonticola

Acid from dulcitolCarbon source utilization tests: dulcitol, 3-phenylpropionate


tests can help to sort out these “atypical” strains:after examining more than 5,000 Serratia strains,we have found no S. liquefaciens, no S. grimesii,and no S. plymuthica strains that were able togrow on adonitol and erythritol and no S.rubidaea strains able to grow on sorbitol as solecarbon source.

Identification of Serratia Biogroups and Biotypes

Biotypes in Serratia marcescens, S. proteamacu-lans, S. plymuthica, and S. rubidaea were firstdefined by numerical taxomony (Grimont et al.,1977b). We recognize 18 biotypes in S. marce-

Table 4. Differential characteristics of Serratia species.

aThe liquefaciens group or complex includes S. liquefaciens, S. proteamaculans, and S. grimesii.Without addition of pyrroloquinoline quinone. All species produce gluconate from glucose in the presence of pyrroloquino-line quinone.bSymbols: +, positive for 90% or more strains in 2-day reading; %, negative for 90% or more strains in 4-day reading; d, testused to differentiate biotypes; v, variable reactions; ( ), 4-day reading; ND, not determined.

Characteristic S.m

arce

scen

s

“liq

uefa

cien

sgr

oup”

a

S.pl

ymut

hica

S.ru

bida

ea

S.od

orif

era

S.fic

aria

S.en

tom

ophi

la

S.fo

ntic

ola

Red pigment d % v v % % % %Potato-like odor % % % v + + % %Good growth at 5°C % + + % + + + +Good growth at 40°C + % % v ND ND + NDIndole % % % % + % % %Tetrathionate reduction d + % % % % % +Gas from glucose agar % + v % % % % v#-Xylosidase % % v + + v % +Oxidation of: Glucose to gluconate (without cofactor)b + % v + + + + %

2-Ketogluconate to 2,5-diketogluconate + d % % % % % %Acid from:

Adonitol v % % + (+) + + +L-Arabinose % + + + + + % +Lactose d v (+) + (+) v % +D-Melibiose % + + + + + % +D-Raffinose d + + + d + %+L-Rhamnose % % % % + + % vD-Sorbitol + + d % + + % +Sucrose + + + + d + + %D-Xylose % + + + + + d d

Lysine decarboxylase + + % d + % % +Ornithine decarboxylase + + % % d % % +Arginine decarboxylase % d % % % % % %Tween 80 hydrolysis + + + + % + + +Carbon source utilization:

Adonitol + d % + + + + +L-Arabinose % + + + + + % +D-Arabitol % % % + % + d +L-Arabitol + % % % + + d +Betaine % % v + % % % %Dulcitol % % % % % % % +meso-Erythritol d % % + d + %+Maltitol % + + + % + % vMelezitose % + + d % + % %D-Melibiose % + + + + + % +Palatinose % + + + % + % +Quinate d % + (+) % + d %L-Rhamnose % d % % + + % +D-Sorbitol + + v % + + % +Sucrose + + + + d + + %D-Tartrate % % % d d % % vTricarballylate % % % d % % % vTrigonelline d % % + + v % %


scens, 4 in S. proteamaculans, 2 in S. grimesii, 3in S. rubidaea, 2 in S. odorifera, and 2 in S. ento-mophila. Characteristics of these biotypes aregiven in Tables 6–9. Carbon source utilizationtests are major elements in the identification ofthese biotypes. Multipoint inoculation devices(e.g., Denley Multipoint Inoculator, DenleyInstruments, Ltd., Bolney Cross, Bolney, Sussex,England), are the most useful for biotyping 20 ormore strains at a time. Biotype identification isfaster (1–4 days) when API strips containing car-bon sources are used. Alternatively, filter paperdisks impregnated with each carbon source canbe deposited on an inoculated minimal agar

(Sifuentes-Osornio and Gröschel, 1987). Differ-entiation between S. proteamaculans biotypesCla and Clb is superfluous; electrophoresis ofproteinases (Grimont et al., 1977a) could not dis-criminate between these two biotypes.

Serotyping of Serratia marcescensThe systematic inventory of somatic (O) antigensof S. marcescens began in 1957 (Davis andWoodward, 1957). Flagellar antigens (H) weredescribed (Ewing et al., 1959) and an antigenicscheme was developed (Ewing, 1986; Ewing etal., 1959; Le Minor and Pigache, 1977, 1978; Le

Table 5. Identification of species in the Serratia liquefaciens complex.a

aSymbols as in Table 4.bThe type strain of S. proteamaculans (ATCC 19323) corresponds to biotype C1c.cThe type strain of S. grimesii (ATCC 14460) corresponds to biotype C1d.The potato-like odor produced by S. odorifera, S. ficaria, and a few strains of S. rubidaea, was identified as a pyrazine(3-isopropyl-2-methoxy-5-methylpyrazine) (Gallois and Grimont, 1985).

CharacteristicS. liquefaciens

Clab

S. proteamaculans

S. grimesiisubsp. proteamaculans subsp. quinovoraRQC1cb EB RB C1dc ADC

Growth on:trans-Aconitate % + + % + % +Adonitol % % (+) % % % %Benzoate % % (+) v % + %m-Erythritol % % + % % % %Gentisate % % % + % % %D-Malate + % % % v (v) (v)L-Rhamnose % % % + d % %m-Tartrate + % % % d % %

Arginine decarboxylase % % % % % + +Esculine hydrolyzed + + + + % + +

Table 6. Identification of Serratia marcescens biotypes.a

aSymbols as in Table 4.bGrowth on D-malate and meso-tartrate is correlated with growth on DL-carnitine (Grimont, unpublished observations).cLactose-positive strains are occasionally isolated.dNonpigmented strains are occasionally isolated.

Characteristic

Biotype

A1 A2 A6 A3 A4

A5

A8

TCT TC TTa b a b a a b c d a b a b c

Growth on:m-Erythritol + + + + + + + + + + + % % % % % % %Trigonelline % % % + % % + % + % % + + + + + % +Quinate and/or 4-hydroxybenzoate % % % % + % % % % + % + + + + % % %3-Hydroxybenzoate % % % % % + + % % % % % % + % % % %Benzoate + + % % % % % % % % % % % % % % % %D-Malate/m-tartrateb + % + + v % % v % + + + v % % + + %Gentisate % % + + + + + v % + + % v + v % % %Lactose % % % % % %c % %c %c %c % %c % %c + %c %c %

Tetrathionate reduction + + + + + + + + + % % + + + + + + +Red pigment +d +d + + +d % % % % % % % % % % % % %


Minor and Sauvageot-Pigache, 1981; Le Minoret al., 1983; Sedlak et al., 1965; Traub, 1981, 1985;Traub and Fukushima, 1979b; Traub and Kleber,1977). The present system consists of 24 somaticantigens (O1 to O24) and 26 flagellar antigens(H1 to H26). Serotyping of S. marcescens is noteasy. Cross-reactions occur between O antigens2 and 3 (the common antigen is referred to asCo/2, 3); 6 and 7; 6, 12, and 14; and 12, 13, and14 (the common antigen is referred to as Co/12,13, 14). Some strains seem to have more than oneO-factor (e.g., O3,21). Cross-reactions betweenO6 and O14 are so extensive that the epi-demiological distinction between these twoantigens (now referred to as O6/O14) wasabandoned.

The accuracy of the O-agglutination test withboiled antigens has been questioned (Gaston etal., 1988; Gaston and Pitt, 1989a). Immunoblot-ting with electrophoresced lipopolysaccharide(LPS) antigens showed that O-agglutining seraoften reacted with a heat-stable surface antigenrather than with LPS. This heat-stable surfaceantigen masks the expression of O-specific LPSantigens. Gaston and Pitt (1989b) proposed asimple dot enzyme immunoassay (dot EIA) thatcould offer greater accuracy than agglutinationtests for serotype identification. Most strainspossess distinct acidic polysaccharides of micro-capsular origin. Oxley and Wilkinson (1988a)have shown that the O13 antigen is a microcap-sular, acid polymer, rather than an integral partof the lipopolysaccharide. No high-molecularweight LPS corresponding to O-side chain mate-rial was detected in O-serogroup reference strainO11 or O13 (Gaston and Pitt, 1989a). The O6/O14 antigen is a partially acetylated acidicglucomannan (Brigden and Wilkinson, 1985).Three different neutral polymers (O-side chainpolysaccharides) have been found in three O14strains. Each of these polymers has been shownto occur also in strains of serogroups O6, O8, orO12. The polymer with a disaccharide repeating-unit of D-ribose and 2-acetamido-2-deoxy-D-galactose present in the O12 reference strain, theO14:H9 reference strain, and in a O13:H7 refer-ence strain all correspond to the Co/12,13,14antigen shared by these strains (Brigden et al.,1985; Brigden and Wilkinson, 1983; Oxley andWilkinson, 1988b).

To overcome the tediousness of measuringflagellar agglutination, an immobilization test insemisolid agar has been described for determina-tion of H antigens (Le Minor and Pigache, 1977).The technique is facilitated by the use of poolsof nonabsorbed sera. No anti-H:9 serum isneeded in pools, as bacteria with H:9 antigen areimmobilized by anti-H:8 and anti-H:10 sera.Once the unknown strain is immobilized by aserum pool, corresponding individual sera are

Table 7. Identification of Serratia rubidaea biotypes.a

aSymbols as in Table 4.bBiotype B1 corresponds to the subspecies designated asS. rubidaea subsp. burdigalensis; B2 to S. rubidaea subsp.rubidaea; and B3 to S. rubidaea subsp. colindalensis.Unpublished observations, F. Grimont and P. A. D. Grimont.

Trait

Biotypeb

B1 B2 B3

Growth on:Histamine v % +D-Melezitose % + +D-Tartrate + % vTricarballylate % v %

Voges-Proskauer (O’Meara) + % vLysine decarboxylase + + %Malonate (Leifson) + + %

Table 8. Identification of Serratia odorifera biotypes.a

aSymbols as in Table 4.bThe type strain corresponds to biotype 1.cSome strains were positive in 3–7 days.

Trait

Biotype

1b 2

Growth on:m-Erythritol % +L-Fucose v +D-Raffinose + %Sucrose + %D-Tartrate + %

Ornithine decarboxylase + %Acid from sucrose + %Acid from raffinose + %c

Table 9. Identification of Serratia entomophila biotypes.a

aSymbols as in Table 4.bThe type strain corresponds to biotype 1.Biotyping of S. marcescens is epidemiologically useful (Gri-mont and Grimont, 1978b, Sifuentes-Osornio et al., 1986).Pigmentation occurs only in five S. marcescens biotypes: A1a,A1b, A2a, A2b, and A6. The 13 other S. marcescens biotypescorrespond to nonpigmented strains: A3a, A3b, A3c, A3d,A4a, A4b, A5, A8a, A8b, A8c, TCT, TT, and TC. Biotypes TTand TC are rarely found and their ecological-epidemiologicalsignificance is unknown. Nonpigmented biotypes A3abcdand A4ab are ubiquitous, whereas nonpigmented biotypesA5, A8abc, and TCT seem restricted to hospitalized patients(these biotypes can however be isolated from sewage-polluted river water). Pigmented biotypes are ubiquitous.

Trait

Biotype

1b 2

Growth on:D-Arabitol + %L-Arabitol % +D-Malate % vQuinate + vD-Xylose % +


used to achieve final identification of H antigens.The following absorbed sera need to be pre-pared: anti-H:8 absorbed by H:10; anti-H:10absorbed by H:8; and anti-H:3 absorbed by H:10.Other cross-reactions are overcome by dilutionof sera. The immobilization test works becauseH antigens are monophasic in S. marcescens.

A total of 175 complete (O+H) serotypes havebeen identified at this time. No plasmid or phagehas yet been found to alter the serotype of astrain.

Some striking correlations between antigeniccomposition and biotype have been shown (Gri-mont et al., 1979a). Pigmented biotypes and non-pigmented biotypes are antigenically segregated.Serotyping subdivides the biotypes, but whentwo biotypes correspond to the same serotype,these biotypes usually differ by only one bio-chemical reaction. Biotypes can be united into“biogroups” in order to clarify biotype-serotypecorrelation (Table 10). These biogroups may bevalid infraspecific taxons.

Serotyping of Other SpeciesAn antigenic scheme for S. ficaria is presentlyavailable (Grimont and Deval, 1982). Foursomatic (O) and one flagellar (H) antigens wereidentified that defined four serovars. All Ameri-can strains studied (isolated from the fig wasp orfrom a human patient) belonged to serotypeO1:H1; strains from the Mediterranean region(Sicily, Tunisia, France) were not so antigenicallyuniform and all four serotypes were found in figsfrom Sicily.

An antigenic scheme for S. rubidaea has alsobeen devised (Grimont et al., manuscript inpreparation). Nine somatic and five flagellarantigens were identified. A total of 53 strainswere distributed among 14 complete (O+H)serotypes.

Bacteriocin Typing of Serratia marcescensTwo different schemes have been developed fortyping Serratia marcescens strains by susceptibil-ity to bacteriocins (“marcescins”). The system ofTraub (1980) was developed since 1971 (Traub etal., 1971) and uses marcescins (incomplete phagestructures) produced by 10 selected strains. In1980, 73 bacteriocin types had been describedand about 93% of strains were typable (Traub,1980). In this system, bacteriocin type 18 is oftenassociated with serotype 06/14:H12 in multiple-drug-resistant nosocomial strains.

Farmer (1972b) independently developedanother system of typing based on susceptibilityto marcescins. After mitomycin induction, 12bacteriocin-producing strains were selected, andthe 12 marcescins could subdivide 93 strains into79 types.

Typing of S. marcescens by production of bac-teriocin was also developed by Farmer (1972a).Twenty-three strains susceptible to bacteriocinswere selected to serve as indicator strains. Eachunknown S. marcescens strain is treated withmitomycin, and the marcescin produced (if any)is identified according to the pattern of suscepti-bility demonstrated by indicator strains. Thismethod could be used to type 91% of the strainsstudied.

Phage Typing of SerratiaFarmer (1975) has developed a phage typing sys-tem common to Serratia marscens, S. liquefa-ciens, and S. rubidaea (marinorubra). With 74phages, 95% of the S. marcescens strains and50% of the S. liquefaciens and S. rubidaea strainscould be typed. Phage typing was found to beconvenient for subdividing prevalent serotypes(Negut et al., 1975). F. Grimont (1977) indepen-dently developed another phage typing system

Table 10. Correlation between biogroups and serotypes in Serratia marcescens.

aBiogroup A1 is composed of biotypes A1a and A1b; A2/6 is composed of A2a, A2b, and A6; A3 of A3a, A3b, A3c, and A3d;A4 of A4a and A4b; A5/8 of A5, A8a, A8b, and A8c; and TCT of TCT and TT.

Biogroupa Serotypes included

A1 O5:H2; O5:H3; O5:H13; O5:H23; O10:H6; and O10:H13A2/6 O5:H23; O6,14:H2; O6,14:H3; O6,14:H8; O6,14:H9; O6,14:H10; O6,14:H13; O8:H3; and O13:H5A3 O3:H5; O3:H11; O4:H5; O4:H18; O5:H6; O5:H15; O6,14:H5; O6,14:H6; O6,14:H16; O6,14:H20; O9:H9;

O9:H11; O9:H15; O9:H17; O12:H5; O12:H9; O12:H10; O12:H11; O12:H15; O12:H16; O12:H17; O12:H18; O12:H20; O12:H26; O13:H11; O13:H17; O15:H3; O15:H5; O15:H8; O15:H9; O17:H4; O18:H21; O18:H26; O22:H11; and O23:H19

A4 O1:H1; O1:H4; O2:H1; O2:H8; O3:H1; O4:H1; O4:H4; O5:H1; O5:H6; O5:H8; O5:H24; O9:H1; O13:H1; O13:H11; and O13:H13

A5/8 O2:H4; O3:H12; O3,21:H12; O4:H12; O5:H4; O6,14:H4; O6,14:H12; O8:H4; O8:H12; O15:H12; and O21:H12TCT O1:H7; O2:H7; O4:H7; O5:H7; O5:H19; O7:H7; O7:H23; O10:H8; O10:H9; O11:H4; O13:H7; O13:H12;

O16:H19; O18:H9; O18:H16; O18:H19; O19:H14; O19:H19; O20:H12; and O24:H6


for nosocomial strains isolated in Bordeaux.Most of these S. marcescens strains gave phagesusceptibility patterns that corresponded tophage types described by F. Grimont (1977);however, strains isolated in other cities of Francewere less often typable, and strains receivedfrom other countries were rarely typable in thissystem.

Typing by Enzyme ElectrophoresisBy using agar gel electrophoresis, Grimont et al.1977a and Grimont and Grimont (1978c)detected seven different proteinases (called P5,P6, P7, P9a, P9b, P11, and P12) produced by strainsof S. marcescens. Each S. marcescens strain canproduce one to four proteinases; 651 strains of S.marcescens, isolated over a period of six years ina French hospital, gave 33 different proteinasepatterns (called zymotypes). However, six zymo-types alone accounted for 76% of the isolates.Proteinase electrophoresis was thus used as anepidemiological marker (Grimont and Grimont,1978c).

Goullet (1978, 1981) showed that the Serratiaspecies were characterized by distinct electro-phoretic patterns of their esterases. However,these esterase profiles were not used as epidemi-ological marker.

Enzyme polymorphism among 99 S. marce-scens isolates was determined by electrophoresisof nine enzymes (Gargallo-Viola, 1989). Morethan one electromorph was found for seven ofthese enzymes (alanine dehydrogenase, catalase,NADP-dependent malic enzyme, 6-phosphoglu-conate dehydrogenase, glucose-6-phosphatedehydrogenase, NADP-dependent glu- tamatedehydrogenase, and indophenol oxidase), and 33distinctive electrophoretic types. One group wasrepresented exclusively by isolates belonging tononpigmented biotypes recovered almostentirely (97%) from clinical samples. The othergroup comprised all isolates belonging to a pig-mented biotype. Electrophoresis of glucose-6-phosphate dehydrogenase alone can distinguishS. marcescens strains belonging to a pigmentedbiotype from those belonging to a nonpigmentedbiotype (Gargallo et al., 1987). A very goodcorrelation was found between electrophoreticperiplasmic protein patterns, enzyme electro-phoresis, and biotyping by carbon source utiliza-tion (Gargallo-Viola and Lopez, 1990).

Restriction PatternsElectrophoretic patterns of fragments producedafter cleavage of total DNA by restriction endo-nucleases have been used to compare S. marce-scens strains in epidemiological studies (McGeeret al., 1990).

A new, generally applicable typing method hasrecently been proposed by Grimont and Gri-mont (1986) in which the DNA restriction frag-ments carrying rRNA genes are visualized afterhybridization with a labelled E. coli 16+23SrRNA probe (rRNA gene restriction patterns).Five to seven DNA restriction fragments (afterdigestion by Bam HI) or 11 to 13 fragments(after digestion by EcoRI) were observed withall Serratia strains tested (Grimont and Grimont,manuscript in preparation). The different S.marcescens biogroups display different rRNAgene restriction pattern. This method should beuseful in resolving occasionally uncertain (orconflicting) results given by serotyping andbiotyping.

Literature Cited

Ackerman, L. J., Kishimoto, R. A., Emerson, J. S. 1971. Non-pigmented Serratia marcescens arthritis in a Teju (Tupi-nambis tequixin). American Journal of VeterinaryResearch 32:823–826.

Alford, L. R., Holmes, N. E., Scott, W. T., Vickery, J. R. 1950.Studies on the preservation of shell eggs. AustralianJournal of Applied Science 1:208–214.

Altemeier, W. A., Culbertson, W. R., Fullen, W. D., McDon-ough, J. J. 1969. Serratia marcescens septicemia. A newthreat in surgery. Archives of Surgery 99:232–238.

Barnum, D. A., Thackeray, E. L., Fish, N. A. 1958. An out-break of mastitis caused by Serratia marcescens. Cana-dian Journal of Comparative and Medical VeterinaryScience 22:392–395.

Bascomb, S., Lapage, S. P., Willcox, W. R., Curtis, M. A. 1971.Numerical classification of the tribe Klebsielleae. Jour-nal of General Microbiology 66:279–295.

Bergan, T., Grimont, P. A. D., Grimont, F. 1983. Fatty acidsof Serratia determined by gas chromatography. CurrentMicrobiology 8:7–11.

Bergey, D. H., Harrison, F. C., Breed, R. S., Hammer, B. W.,Huntoon, F. M. 1923. Bergey’s manual of determinativebacteriology. Baltimore, Williams & Wilkins.

Berkowitz, D. M., Lee, W. S. 1973. A selective medium forisolation and identification of Serratia marcescens.Abstracts of the Annual Meeting of the American Soci-ety for Microbiology 1973:105.

Bizio, B. 1823. Lettera di Bartolomeo Bizio al chiarissimocanonico Angelo Bellani sopra il fenomeno dellapolenta porporina. Biblioteca Italiana o sia Giornale diLetteratura Scienze a Arti 30:275–295.

Boam, G. W., Sanger, V. L., Cowan, D. F., Vaughan, D. P.1970. Subcutaneous abscesses in iguanid lizards. Journalof the American Veterinary Medical Association157:617–619.

Bollet, C., Gulian, C., Mallet, M. N., Estrangin, E., de Micco,P. 1989. Isolation of Serratia rubidaea (Stapp) from freshand spoiled coconuts. Tropical Agriculture 66:342–345.

Bouvet, O. M. M., Grimont, P. A. D., Richard, C., Aldová,E., Hausner, O., Gabrhelová, M. 1985. Budvicia aquaticagen. nov., sp. nov.: a hydrogen sulfide-producing memberof the Enterobacteriaceae International Journal of Sys-tematic Bacteriology. 35:60–64.


Bouvet, O. M. M., Lenormand, P., Grimont, P. A. D. 1989.Taxonomic diversity of the D-glucose oxidation pathwayin the Enterobacteriaceae. International Journal of Sys-tematic Bacteriology 39:61–67.

Braun, V., Günter, H., Nuess, B., Tautz, C. 1985. Hemolyticactivity of Serratia marcescens. Archives of Microbiol-ogy. 141:371–376.

Breed, R. S., Breed, M. E. 1924. The type species of the genusSerratia, commonly known as Bacillus prodigiosus. Jour-nal of Bacteriology 9:545–557.

Breed, R. S., Breed, M. E. 1927. The genus Serratia Bizio.Zentralblatt für Bakteriologie, Parasitenkunde undInfektionskrankheiten, Abt. 2, 71:435–440.

Breed, R. S., Murray, E. G. D., Hitchens, A. P. 1948. Bergey’smanual of determinative bacteriology. 6th ed.Baltimore, Williams & Wilkins.

Breed, R. S., Murray, E. G. D., Smith, N. R. 1957. Bergey’smanual of determinative bacteriology. 7th ed. Balti-more, Williams & Wilkins.

Brigden, C. J., Furn, S., Wilkinson, S. G. 1985. Structuralstudies of neutral polymers isolated from thelipopolysaccharides of Serratia marcescens O6 (strain C.D. C. 862-57) and O12 (C.D.C. 6320-58). CarbohydrateResearch 139:298–301.

Brigden, C. J., Wilkinson, S. G. 1983. Lipopolysaccharidefrom the O14 type strain of Serratia marcescens: struc-tural studies of a polymeric fraction. CarbohydrateResearch 115:183–1901.

Brigden, C. J., Wilkinson, S. G. 1985. Structural studies ofacidic glucomannans from strains of Serratia marcescensO14 and O6. Carbohydrate Research 138:267–276.

Brisou, J., Cadeillan, J. 1959. Etude sur les Serratia. Bulletinde l’Association des Diplomés de Microbiologie de laFaculté de Pharmacie de Nancy. 75:34–39.

Brisou, B., Richard, C., Lenriot, A. 1972. Intérêt tax-onomique de la recherche de la #-xylosidase chezles Enterobacteriaceae. Annales de l’Institut Pasteur123:341–347.

Bucher, G. E. 1959. Bacteria of grasshoppers of westernCanada. Journal of Insect Pathology 1:391–405.

Bucher, G. E. 1960. Potential bacterial pathogens of insectsand their characteristics. Journal of Insect Pathology2:172–195.

Bucher, G. E. 1963a. Nonsporulating bacterial pathogens.117–147. Steinhaus, E. A. (ed.) Insect pathology. Anadvanced treatise, vol. 2. New York. Academic Press.

Bucher, G. E. 1963b. Transmission of bacterial pathogens bythe ovipositor of a hymenopterous parasite. Journal ofInsect Pathology 5:277–283.

Bucher, G. E., Stephens, J. M. 1959. Bacteria of grasshoppersof western Canada. Journal of Insect Pathology 1:356–373.

Burnside, C. E. 1928. A septicemic condition of adult bees.Journal of Economic Entomology 21:379–386.

Cabrera, H. A. 1969. An outbreak of Serratia marcescens andits control. Archives of Internal Medicine 123:650–655.

Caldis, P. D. 1927. Etiology and transmission of endosepsis(internal rot) of the fruit of the fig. Hilgardia 2:287–328.

Capponi, M., Sureau, P., Le Minor, L. 1956. Contribution ál’étude des salmonelles du Centre-Vietnam. Bulletin dela Société de Pathologie Exotique. 49:796–801.

Carter, G. R. 1973. Diagnostic procedures in veterinarymicrobiology. Springfield, Illinois. Charles C. Thomas..

Cate, J. C. 1972. Isolation of Serratia marcescens from stoolswith an antibiotic plate. 763–764. Hejzlar, M., Semonsky,M. and Masák, S. (ed.) Advances in antimicrobial and

antineoplastic chemotherapy. Progress in research andclinical application. Proceedings of the 7th InternationalCongress of Chemotherapy, vol. 1/2: Munich, Berlin,Vienna, Urban & Schwarzenberg.

Colwell, R. R., Mandel, M. 1965. Adansonian analysis anddeoxyribonucleic acid base composition of Serratiamarcescens. Journal of Bacteriology 89:454–461.

Combe, E. 1933. Etude comparative de trois bactéries chro-mogènes à pigment rouge: Serratia marcescens Bizio(Bacillus prodigiosus Flügge), Serratia kiliensis ComitéS.B.A. (Bacterium h Breunig), Serratia esseyana n. sp.Combe (Bacille rouge d’Essey Lasseur). University ofNancy. Nancy, France.

Daschner, F. D. 1980. The epidemiology of Serratia marce-scens. 187–196. von Graevenitz, A., and Rubin, S. J. (ed.)The genus Serratia. Boca Raton, CRC Press.

Dauenhauer, S. A., Hull, R. A., Williams, R. P. 1984. Cloningand expression in Escherichia coli of Serratia marcescensgenes encoding prodigiosin biosynthesis. Journal of Bac-teriology 158:1128–1132.

Davis, B. R., Woodward, J. M. 1957. Some relationships ofthe somatic antigens of a group of Serratia marce-scens cultures. Canadian Journal of Microbiology3:591–597.

Denis, F. A., Blanchard, P. 1975. Enquête sur les porteursintestinaux de Serratia. Nouvelle Presse Médicale4:2114–2115.

Deom, J., Mortelmans, J. 1953. Etude d’une souchepathogène de Serratia marcescens. Revue d’Immunolo-gie 17:394–398.

Ding, M. J., Williams, R. P. 1983. Biosynthesis of prodigiosinby white strains of Serratia marcescens isolated frompatients. Journal of Clinical Microbiology 17:476–480.

Duran-Reynals, F., Clausen, H. J. 1937. A contagious tumor-like condition in the lizard (Anolis equestris) as inducedby a new bacterial species, Serratia anolium (sp. n.).Journal of Bacteriology 33:369–379.

Ehrenberg, C. G. 1848. Communication presented at the ple-nary session of the Royal Prussian Academy of Sciences,October 26. 346–362. Akademie der Wissenschaften zuBerlin.

Eisenberg, J. 1886. Bakteriologische Diagnostik. Hülfs-Tabellen beim Praktischen Arbeiten. Hamburg,Leopold Voss.

El Sanoussi, S. M., El Sarag, M. S. A., Mohamed, S. E. 1987.Properties of Serratia marcescens isolated from diseasedhoneybee (Apis mellifera) larvae. Journal of GeneralMicrobiology 133:215–219.

Ewing, W. H. 1963. An outline of nomenclature for the familyEnterobacteriaceae. International Bulletin of Bacterio-logical Nomenclature and Taxonomy 13:95–110.

Ewing, W. H. 1986. Edwards and Ewing’s Identification ofEnterobacteriaceae. 4th ed. Elsevier. New York.

Ewing, W. H., Davis, B. R., Fife, M. A. 1972. Biochemicalcharacterization of Serratia liquefaciens and Serratiarubidaea. Atlanta, Center for Disease Control.

Ewing, W. H., Davis, B. R., Johnson, J. G. 1962. The genusSerratia: Its taxonomy and nomenclature. InternationalBulletin of Bacteriological Nomenclature and Taxon-omy 12:47–52.

Ewing, W. H., Davis, B. R., Reavis, R. W. 1959. Studies onthe Serratia group. Atlanta, Center for Disease Control.

Ewing, W. H., Davis, B. R., Fife, M. A., Lessel, E. F. 1973.Biochemical characterization of Serratia liquefaciens(Grimes and Hennerty) Bascomb et al. (formerly Enter-obacter liquefaciens) and Serratia rubidea (Stapp) comb.


nov. and designation of type and neotype strains. Inter-national Journal of Systematic Bacteriology 23:217–225.

Farmer, J. J., III. 1972a. Epidemiological differentiation ofSerratia marcescens: Typing by bacteriocin production.Applied Microbiology 23:218–225.

Farmer, J. J., III. 1972b. Epidemiological differentiation ofSerratia marcescens: Typing by bacteriocin sensitivity.Applied Microbiology 23:226–231.

Farmer, J. J., III. 1975. Lysotypie de Serratia marcescens.Archives Roumaines de Pathologie Experimentale et deMicrobiologie 34:189.

Farmer, J. J., III. Davis, B. R., Hickman-Brenner, F. W.,McWhorter, A., Huntley-Carter, G. P., Asbury, M. A.,Riddle, C., Wathen-Grady, H. G., Elias, C., Fanning, G.R., Steigerwalt, A. G., O’Hara, C. M., Morris, G. K.,Smith, P. B., Brenner, D. J. 1985. Biochemical identifica-tion of new species and biogroups of Enterobacteriaceaeisolated from clinical specimens. Journal of ClinicalMicrobiology 21:46–76.

Farmer, J. J., III. Davis, B. R., Hickman, F. H., Presley, D. B.,Bodey, G. P., Negut, M., Bobo, R. A. 1976. Detection ofSerratia outbreaks in hospital. Lancet ii:455–459.

Farmer, J. J., III. Silva, F., Williams, D. R. 1973. Isolation ofSerratia marcescens on deoxyribonuclease- toluidineblue-cephalothin agar. Applied Microbiology 25:151–152.

Farrar, W. E. Jr. 1980. Antimicrobial susceptibility of clinicalisolates, synergistic effects, and beta-lactamases of Ser-ratia. 121–138.von Graevenitz, A., and Rubin, S. J. (ed.)The genus Serratia. Boca Raton, CRC Press.

Feistner, G., Korth, H., Ko, H., Pulverer, G., Budzikiewicz,H. 1983. Ferrorosamine a from Erwinia rhapontici. Cur-rent Microbiology 8:239–243.

Fortineau, L. 1904. Erythrobacillus pyosepticus et bactériesrouges. M. D. Thesis. Faculté de Médecine. Paris,

Franczek, S. P., Williams, R. P., Hull, S. I. 1986. A survey ofpotential virulence factors in clinical and environmentalisolates of Serratia marcescens. Journal of MedicalMicrobiology 22:151–156.

Fuchs, R. L., McPherson, S. A., Drahos, D. J. 1986. Cloningof a Serratia marcescens gene encoding chitinase.Applied and Environmental Microbiology 51:504–509.

Fulton, M., Forney, C. E., Leifson, E. 1959. Identification ofSerratia occurring in man and animals. Canadian Journalof Microbiology 5:269–275.

Gallois, A., Grimont, P. A. D. 1985. Pyrazines responsible forthe potatolike odor produced by some Serratia andCedecea strains. Applied and Environmental Microbiol-ogy 50:1048–1051.

Gargallo, D., Lorèn, J. G., Guinea, J., Viñas, M. 1987. Glu-cose-6-phosphate dehydrogenase alloenzymes and theirrelationship to pigmentation in Serratia marcescens.Applied and Environmental Microbiology 53:1983–1986.

Gargallo-Viola, D. 1989. Enzyme polymorphism, prodigiosinproduction, and plasmid fingerprints in clinical and nat-urally occuring isolates of Serratia marcescens. Journalof Clinical Microbiology 27:860–868.

Gargallo-Viola, D., Lopez, D. 1990. Numerical analysis ofelectrophoretic periplasmic protein patterns, a possiblemarker system for epidemiologic studies. Journal ofClinical Microbiology 28:136–139.

Gaston, M. A., Duff, P. S., Pitt, T. L. 1988. Lipopolysaccharideheterogeneity in strains of Serratia marcescens aggluti-nated by O14 antiserum. Current Microbiology 17:27–31.

Gaston, M. A., Pitt, T. 1989a. O-Antigen specificities of theserotype strains of Serratia marcescens. Journal of Clin-ical Microbiology 27:2697–2701.

Gaston, L. A., Pitt, T. 1989b. Improved O-serotyping methodfor Serratia marcescens. Journal of Clinical Microbiology27:2702–2705.

Gaughran, E. R. L. 1969. From superstition to science: Thehistory of a bacterium. Transactions of the New YorkAcademy of Sciences 31:3–24.

Gavini, F., Ferragut, C., Izard, D., Trinel, P. A., Leclerc, H.,Lefebvre, B., Mossel, D. A. A. 1979. Serratia fonticola, anew species from water. International Journal of Sys-tematic Bacteriology 29:92–101.

Gavini, F., Izard, D., Leclerc, H., Desmonceaux, M., Gayral,J. P. 1980. Carbon source assimilation tests: comparisonbetween a conventional method and a microtechnic(API), in study of Enterobacteriaceae. Zbl. Bakt., I.Abt. Orig. C1:182–187.

Giammanco, G., Amato, R. 1982. Sulla presenza in Sicilia diuna nuova specie di “Serratia” di recente descrizione.L’Igiene Moderna 77:627–632.

Gill, V. J., Farmer III, J. J., Grimont, P. A. D., Asbury, M. A.,McIntosh, C. L. 1981. Serratia ficaria isolated fromhuman clinical specimen. Journal of Clinical Microbiol-ogy 14:234–236.

Goldin, M., Shaffer, J. G., Brown, E. 1969. A new selectiveand differential medium for Serratia. BacteriologicalProceedings 1969:96.

Gosden, P. E., Ware, G. C. 1976. The aerobic bacterial floraof the anal sac of the red fox. Journal of Applied Bacte-riology 41:271–275.

Goullet, P. 1978. Characterization of Serratia marcescens, S.liquefaciens, S. plymuthica, and S. marinorubra by theelectrophoresis patterns of their esterases. Journal ofGeneral Microbiology 108:275–281.

Goullet, P. 1981. Characterization of Serratia odorifera S.fonticola and S. ficaria by the electrophoresis patterns oftheir esterases. Journal of General Microbiology127:161–167.

Greenup, P., Blazevic, D. J. 1971. Antibiotic susceptibilitiesof Serratia marcescens and Enterobacter liquefaciens.Applied Microbiology 22:309–314.

Grimes, M. 1961. Classification of the Klebsiella-Aerobactergroup with special reference to the cold tolerant meso-philic Aerobacter types. Intern. Bull. Bact. Nomen.Taxon. 11:111–129.

Grimes, M., Hennerty, A. J. 1931. A study of bacteria belong-ing to the sub-genus Aerobacter. Scientific Proceedingsof the Royal Dublin Society, New Series. 20:89–97.

Grimont, F. 1977. Les bactériophages des Serratia etbactéries voisines. Taxonomie et lysotypie. Thèse dePharmacie. University of Bordeaux II. Bordeaux,France.

Grimont, F., Grimont, P. A. D. 1986. Ribosomal ribonucleicacid gene restriction patterns as potential taxonomictools. Annales de l’Institut Pasteur, Microbiologie.137B:165–175.

Grimont, P. A. D. 1977. Le Genre Serratia. Taxonomie etapproche écologique. Ph.D. Thesis. University of Bor-deaux I. Bordeaux, France.

Grimont, P. A. D., Ageron, E. 1989. Enterobacter canceroge-nus (Urosevic, 1966) Dickey and Zumoff 1988, a seniorsubjective synonym of Enterobacter taylorae Farmer etal. (1985). Research in Microbiology 140:459–465.

Grimont, P. A. D., Deval, C. 1982. Somatic and flagellarantigens of Serratia ficaria from the United States and


the Mediterranean region. Current Microbiology 7:363–366.

Grimont, P. A. D., Dulong de Rosnay, H., L. C. 1972. Numer-ical study of 60 strains of Serratia. Journal of GeneralMicrobiology 72:259–268.

Grimont, P. A. D., Grimont, F. 1978a. The genus Serratia.Annual Review of Microbiology 32:221–248.

Grimont, P. A. D., Grimont, F. 1978b. Biotyping of Serratiamarcescens and its use in epidemiological studies. Jour-nal of Clinical Microbiology 8:73–83.

Grimont, P. A. D., Grimont, F. 1978c. Proteinase zymogramsof Serratia marcescens as an epidemiological tool. Cur-rent Microbiology 1:15–18.

Grimont, P. A. D., Grimont, F. 1981. The genus Serratia p.1187–1203. In: Starr, M. P., Stolp, H., Truper, H. G.,Balows, A., and Schlegel, H. G. (ed.), The prokaryotes:A handbook on habitats, isolation, and identification ofbacteria. Springer-Verlag. Berlin,

Grimont, P. A. D., Grimont, F. 1984. Genus Serratia Bizio1823, 288AL, p. 477–484. In: Krieg, N. R., and Holt, J. G.(ed.), Bergey’s manual of systematic bacteriology, vol. 2.Williams and Wilkins. Baltimore,

Grimont, P. A. D., Grimont, F., Dulong de Rosnay, H. L. C.1977a. Characterization of Serratia marcescens, S. lique-faciens, S. plymuthica, and S. marinorubra by electro-phoresis of their proteinases. Journal of GeneralMicrobiology 99:301–310.

Grimont, P. A. D., Grimont, F., Dulong de Rosnay, H. L. C.,Sneath, P. H. A. 1977b. Taxonomy of the genus Serratia.Journal of General Microbiology 98:39–66.

Grimont, P. A. D., Grimont, F., Irino, K. 1982a. Biochemicalcharacterization of Serratia liquefaciens sensu stricto,Serratia proteamaculans, and Serratia grimesii sp. nov.Current Microbiology 7:69–74.

Grimont, P. A. D., Grimont, F., Le Minor, S., Davis, B.,Pigache, F. 1979a. Compatible results obtained from bio-typing and serotyping in Serratia marcescens. Journal ofClinical Microbiology 10:425–432.

Grimont, P. A. D., Grimont, F., Lysenko, O. 1979b. Speciesand biotype identification of Serratia strains associatedwith insects. Current Microbiology 2:139–142.

Grimont, P. A. D., Grimont, F., Richard, C., Davis, B. R.,Steigerwalt, A. G., Brenner, D. J. 1978a. Deoxyribo-nucleic acid relatedness between Serratia plymuthicaand other Serratia species with a description ofSerratia odorifera sp. nov. (type strain: ICPB 3995).International Journal of Systematic Bacteriology28:453–463.

Grimont, P. A. D., Grimont, F., Starr, M. P. 1978b. Serratiaproteamaculans (Paine and Stansfield) comb. nov., asenior subjective synonym of Serratia liquefaciens(Grimes and Hennerty) Bascomb et al. InternationalJournal of Systematic Bacteriology 28:503–510.

Grimont, P. A. D., Grimont, F., Starr, M. P. 1979c. Serratiaficaria sp. nov. a bacterial species associated withSmyrna figs and the fig wasp Blastophaga psenes. Cur-rent Microbiology 2:277–282.

Grimont, P. A. D., Grimont, F., Starr, M. P. 1981. Serratiaspecies isolated from plants. Current Microbiology5:317–322.

Grimont, P. A. D., Irino, K., Grimont, F. 1982b. The Serratialiquefaciens-S. proteamaculans-S. grimesii complex:DNA relatedness. Current Microbiology 7:63–68.

Grimont, P. A. D., Jackson, T. A., Ageron, E., Noonan, M. J.1988. Serratia entomophila sp. nov. associated withamber disease in the New Zealand grass grub Costelytra

zealandica. International Journal of Systematic Bacteri-ology 38:1–6.

Groscop, J. A., Brent, M. M. 1964. The effects of selectedstrains of pigmented microorganisms on small free-living amoebae. Canadian Journal of Microbiology10:579–584.

Harpster, M. H., Dunsmuir, P. 1989. Nucleotide sequence ofthe chitinase B gene of Serratia marcescens QMB1466.Nucleic Acids Research 17:5395.

Harrison, F. C. 1924. The “miraculous” microorganism.Transactions of the Royal Society of Canada. 18:1–17.

Hechler, U., Van Den Weghe, M., Martin, H. H., Frére, J.-M.1989. Overproduced #-lactamase and the outer-membrane barrier as resistance factors in Serratiamarcescens highly resistant to #-lactamase-stable #-lactam antibiotics. Journal of General Microbiology135:1275–1290.

Hedges, R. W. 1980. R factors of Serratia. 139–156.von Grae-venitz, A., and Rubin, S. J. (ed.) The genus Serratia. BocaRaton, CRC Press.

Hefferan, M. 1904. A comparative and experimental study ofbacilli producing red pigment. Zentralblatt für Bakteri-ologie, Parasitenkunde, und Infektionskrankheiten,Abt. 2 11:311–317, 397–404, 456–475, 520–540.

Ismail, G., Som, F. M. 1982. Hemagglutination reaction andepithelial cell adherence activity of Serratia marcescens.J. Gen. Appl. Microbiol. 28:161–168.

Iverson, K. L., Bromel, M. C., Anderson, A. W., Freeman,T. P. 1984. Bacterial symbionts in the sugar beet rootmaggot, Tetanops myopaeformis (von Röder). Appliedand Environmental Microbiology 47:22–27.

Izawa, H., Nagabayashi, T., Kazuno, Y., Soekawa, M. 1971.Occurrence of death of chick embryos by Serratia marce-scens infection. Japanese Journal of Bacteriology26:200–204.

Jackson, C. G. Jr., Fulton, M. 1976. A turtle colony epizooticapparently of microbial origin. Journal of Wildlife Dis-ease 6:466–468.

Jackson, T. A., Pearson, J. F. 1986. Control of the grassgrub, Costelytra zealandica (White) (Coleoptera: Scara-baeidae), by application of the bacteria Serratia spp.causing honey disease. Bulletin of EntomologicalResearch 76:69–76.

Janota-Bassalik, L. 1963. Psychrophiles in low-moor peat.Acta Microbiologica Polonica 12:25–40.

Jessop, H. L., Lambert, P. A. 1985. Immunochemical charac-terization of the outer membrane complex of Serratiamarcescens and identification of the antigens accessibleto antibodies on the cell surface. Journal of GeneralMicrobiology 131:2343–2348.

Jones, J. D., Grady, K. L., Suslow, T. V., Bedbrook, J. R. 1986.Isolation and characterization of genes encoding twochitinase enzymes from Serratia marcescens. EMBO J.5:467–473.

Joshi, S., Kozlowski, M., Richens, S., Comberbach, D. M.1989. Chitinase and chitobiase production during fer-mentation of genetically improved Serratia liquefaciens.Enzyme Microb. Technol. 11:289–296.

Kamata, R., Yamamoto, T., Matsumoto, K., Maeda, H. 1985.A Serratial protease causes vascular permeability reac-tion by activation of the Hageman factor-dependentpathway in guinea pigs. Infection and Immunity 48:747–753.

Kaska, M. 1976. The toxicity of extracellular proteases of thebacterium Serratia marcescens for larvae of greater wax


moth, Galleria mellonella. Journal of InvertebratePathology 27:271.

Kisumi, M., Nakanishi, N., Takagi, T., Chibata, I. 1977.L-Histidine production by histidase-less regulatorymutants of Serratia marcescens constructed by transduc-tion. Applied and Environmental Microbiology 34:465–472.

Kisumi, M., Takagi, T., Chibata, I. 1978. Construction of anL-Arginine-producing mutant in Serratia marcescens:use of the wide substrate specificity of acetylornithinase.Journal of Biochemistry Tokyo, 84:881–890.

Komatsubara, S., Kisumi, M., Chibata, I. 1980. Transduc-tional construction of an isoleucine-producing strain ofSerratia marcescens. Journal of General Microbiology119:51–61.

Komatsubara, S., Kisumi, M., Chibata, I. 1983. Transduc-tional construction of a threonine-hyperproducing strainof Serratia marcescens: lack of feedback controls of threeaspartokinases and two homoserine dehydrogenases.Applied and Environmental Microbiology 45:1445–1452.

Kreger, A. S., Griffin, O. K. 1975. Cornea-damaging proteasesof Serratia marcescens. Invest. Ophthalmol. 14:190.

Kushner, D. J., Harvey, G. T. 1962. Antibacterial substancesin leaves: Their possible role in insect resistance to dis-ease. Journal of Insect Pathology 4:155–184.

Lahellec, C., Meurier, C., Bennejean, G., Catsaras, M. 1975.A study of 5920 strains of psychrotrophic bacteria iso-lated from chickens. Journal of Applied Bacteriology38:89–97.

Lakso, J. U., Starr, M. P. 1970. Comparative injuriousness toplants of Erwinia spp. and other enterobacteria fromplants and animals. Journal of Applied Bacteriology33:692–707.

Legakis, N. J., Papavassiliou, J. Th., Xilinas, M. E. 1976. Inos-itol as a selective substrate for the growth of klebsiellaeand serratiae. Zentralblatt für Bakteriologie, Parasiten-kunde, Infektionskrankheiten, und Hygiene, Abt. 1Orig., Reihe A 235:453–458.

Lehmann, K. B., Neumann, R. 1896. Atlas und Grundriss derBakteriologie und Lehrbuch der speciellen bakteriolo-gischen Diagnostik, Teil II. Munich, Lehmann.

Le Minor, S., Benazet, F., Martin, L. 1983. Nouveaux facteursantigéniques O (023) et H (H26) de Serratia marcescens.Annales de Microbiologie, (Institut Pasteur). 134 B:447–449.

Le Minor, L., Chippaux, M., Pichinoty, F., Coynault, C.,Piéchaud, M. 1970. Méthodes simples permettant derechercher la tétrathionate-réductase en cultures liq-uides on sur colonies isolées. Annales de l’InstitutPasteur. 119:733–737.

Le Minor, S., Pigache, F. 1977. Étude antigénique de souchesde Serratia marcescens isolées en France. Annales deMicrobiologie 128B:207–214.

Le Minor, S., Pigache, F. 1978. Etude antigénique de souchesde Serratia marcescens isolées en France. Annales deMicrobiologie 129B:407–423.

Le Minor, S., Sauvageot-Pigache, F. 1981. Nouveaux facteursantigéniques H (H21–H25) et (021) de Serratia marce-scens: subdivision des facteurs 05, 010, 016. Annales deMicrobiologie (Institut Pasteur). 132 A:239–252.

Liu, P. V. 1961. Observation on the specificities of the extra-cellular antigens of the genera Aeromonas and Serratia.Journal of General Microbiology 24:145.

Loiseau-Marolleau, M. L., Laforest, H. 1976. Contribution àl’étude de la flore bactérienne des aliments en milieu

hospitalier. Médecine et Maladies Infectieuses 6:160–171.

Lyerly, D., Gray, L., Kreger, A. 1981. Characterization ofrabbit corneal damage produced by Serratia keratitisand by a Serratia protease. Infection and Immunity33:927–932.

Lyerly, D., Kreger, A. 1979. Purification and characterizationof Serratia marcescens metalloprotease. Infection andImmunity 24:411–421.

Lyerly, D., Kreger, A. 1983. Importance of serratia proteasein the pathogenesis of experimental of Serratia marce-scens pneumonia. Infection and Immunity 40:113–119.

Lynch, D. L., Kenealy, W. R. 1976. A selective medium forthe isolation of Serratia sp. from raw sewage. MicrobiosLetters 1:35–37.

Lysenko, O. 1961. Pseudomonas—an attempt at a generalclassification. Journal of General Microbiology 25:379–408.

Lysenko, O. 1976. Chitinase of Serratia marcescens and itstoxicity to insects. Journal of Invertebrate Pathology27:385–386.

Maeda, H., Molla, A., Oda, T., Katsuki, T. 1987. Internaliza-tion of serratial protease into cells as an enzyme-inhibi-tor complex with "2 macroglobulin and regeneration ofprotease activity and cytotoxicity. Journal BiologicalChemistry 262:10946–10950.

Maki, D. G., Hennekens, C. G., Phillips, C. W., Shaw, W. V.,Bennet, J. V. 1973. Nosocomial urinary tract infectionwith Serratia marcescens: An epidemiologic study. Jour-nal of Infectious Diseases 128:579–587.

Mandel, M., Rownd, R. 1964. Deoxyribonucleic acid basecomposition in the Enterobacteriaceae: An evolutionarysequence. 585–597. Leone, C. A. (ed.) Taxonomic bio-chemistry and serology. New York. Ronald Press.

Marre, R., Hacker, J., Braun, V. 1989. The cell-bound hemol-ysin of Serratia marcescens contributes to uropathoge-nicity. Microbial Pathogenesis 7:153–156.

Martinec, T., Kocur, M. 1960. The taxonomic status of Serra-tia plymuthica (Lehman and Neumann) Bergey et al.and of Serratia indica (Eisenberg) Bergey et al. Interna-tional Bulletin of Bacteriological Nomenclature andTaxonomy 10:247–254.

Martinec, T., Kocur, M. 1961a. The taxonomic status of Ser-ratia marcescens Bizio. International Bulletin of Bacte-riological Nomenclature and Taxonomy 11:7–12.

Martinec, T., Kocur, M. 1961b. A taxonomic study of themembers of the genus Serratia. International Bulletin ofBacteriological Nomenclature and Taxonomy 11:73–78.

Martinec, T., Kocur, M. 1961c. Contribution to the taxonomicstudies of Serratia kiliensis (Lehmann and Neumann)Bergey. International Bulletin of BacteriologicalNomenclature and Taxonomy 11:87–90.

Martinec, T., Kocur, M. 1961d. Taxonomická studie rodu Ser-ratia. Folia Facultatis Scientarum Naturalium Universi-talis Purkynianae Brunensis. 2:1–77.

Martinez, J. L., Cercenado, E., Baquero, E., Pérez-Diaz, J. C.,Delgado-Iribarren, A. 1987. Incidence of aerobactinproduction in Gram-negative hospital isolates. FEMSMicrobiology Letters 43:351–353.

Matsumoto, K., Maeda, H., Takata, K., Kamata, R.,Okamura, R. 1984. Purification and characterization offour proteases from a clinical isolate of Serratia marce-scens kums 3958. Journal of Bacteriology 157:225–232.

Matsumoto, H., Tazaki, T., Hosogaya, S. 1973. A generalizedtransducing phage of Serratia marcescens. JapaneseJournal of Microbiology 17:473–479.


Matsumoto, K., Yamamoto, T., Kamata, R., Maeda, H. 1984.Pathogenesis of Serratial infection: activation of theHageman factor-prekallikein cascade by Serratial pro-tease. Journal of Biochemistry 96:739–749.

Matsuyama, T., Murakami, T., Fujita, M., Fujita, S., Yano, I.1986. Extracellular vesicle formation and biosurfactantproduction by Serratia marcescens. Journal of GeneralMicrobiology 132:865–875.

McCoy, R. H., Scidler, R. J. 1973. Potential pathogens in theenvironment: Isolation, enumeration, and identificationof seven genera of intestinal bacteria associated withsmall green pet turtles. Applied Microbiology 25:534–538.

McGeer, A., Low, D. E., Penner, J., Ng, J., Goldman, C.,Simor, A. E. 1990. Use of molecular typing to study theepidemiology of Serratia marcescens. Journal of ClinicalMicrobiology 28:55–58.

Molla, A., Akaike, T., Maeda, H. 1989. Inactivation of vari-ous proteinase inhibitors and the complement system inhuman plasma by the 56-Kilodalton proteinase. Infec-tion and Immunity 57:1868–1871.

Molla, A., Matsumoto, K., Oyamada, I., Katsuki, T., Maeda,H. 1986. Degradation of protease inhibitors, immuno-globulins, and other serum proteins by Serratia proteaseand its toxicity to fibroblasts in culture. Infection andImmunity 53:522–529.

Müller, H. E., Steigerwalt, A. G., Brenner, D. J. 1986. Isola-tion of Serratia fonticola from birds. Zbl. Bakt. Hyg. A261:212–218.

Negut, M., Davis, B. R., Farmer III, J. J. 1975. Différenciationépidémiologique de Serratia marcescens: Comparaisonentre lysotypie et sérotypie. Archives Roumaines dePathologie Expérimentale et de Microbiologie 34:189.

Ohshima, Y., Ohtomo, T., Yoshida, K. 1984. Encapsulationand mouse-virulence of Serratia marcescens strain SM-1and its variants in relation to colonial morphology. ActaMicrobiologica Hungarica 31:55–60.

Old, D. C., Adegbola, R., Scott, S. S. 1983. Multiple fimbrialhaemagglutinins in Serratia species. Med. Microbiol.Immunol. 172:107–115.

Olexy, V. M., Mucha, D. K., Bird, T. J., Grieble, H. G.,Farrand, S. K. 1982. An R plasmid of Serratia marcescenstransferable to Pseudomonas aeruginosa. Chemother-apy 28:6–17.

O’Meara, R. A. Q. 1931. A simple, delicate and rapid methodof detecting the formation of acetylmethylcarbinol bybacteria fermenting carbohydrate. Journal of Pathologyand Bacteriology 34:401.

Oxley, D., Wilkinson, S. G. 1988a. Studies of lipopolysaccha-rides from two strains (C.D.C. 3607–60 and IP 421) ofSerratia marcescens O13: structure of the putative O13antigen. Carbohydrate Research 172:275–286.

Oxley, D., Wilkinson, S. G. 1988b. Structural studies of glu-corhamnans isolated from the lipopolysaccharides ofreference strains for Serratia marcescens serogroups O4and O7, and of an O14 strain. Carbohydrate Research175:111–117.

Paillot, A. 1916. Existence de plusieurs variétés et races deCoccobacilles dans les septicémies naturelles du Han-neton. Comptes Rendus Hebdomadaires des Séances del’Académie des Sciences. 163:531–534.

Paine, S. G., Berridge, E. M. 1921. Studies in bacteriosis. V.Further investigation of a suggested bacteriolytic actionin Protea cynaroides affected with the leaf-spot dis-ease.Annals of Applied Biology 8:20–26.

Paine, S. G., Stansfield, H. 1919. Studies in bacteriosis.Annals of Applied Biology 6:27–39.

Parès, Y. 1964. Action de Serratia marcescens dans le cyclebiologique des métaux. Annales de l’Institut Pasteur.107:136–141.

Perestelo, F., Falcon, M. A., De La Fuente, G. 1989. Produc-tion of vanillic acid from vanillin by resting cells ofSerratia marcescens. Applied and Environmental Micro-biology 55:1660–1662.

Pesson, P., Toumanoff, C., Haradas, C. 1955. Etude desépizooties bactériennes observées dans les élevagesd’insectes xylophages (Rhyncolus porcatus Germain,Scolytus scolytus Fabricius, Scolytus (Scolytochelus) mul-tistriatus Marsham). Annales des Epiphyties 6:315–328.

Phaff, H. J., Miller, M. W. 1961. A specific microflora associ-ated with the fig wasp, Blastophaga psenes Linnaeus.Journal of Insect Pathology 3:233–243.

Podgwaite, J. D., Cosenza, B. J. 1976. A strain of Serratiamarcescens pathogenic for larvae of Lymantria dispar:Infectivity and mechanisms of pathogenicity. Journal ofInvertebrate Pathology 27:199–208.

Poole, K., Schiebel, E., Braun, V. 1988. Molecular character-ization of the hemolysin determinant of Serratia marce-scens. 170:3177–3188.

Quesenberry, K. E., Short, B. G. 1983. Serratia marcescensinfection in a blue and gold macaw. Journal of the Amer-ican Veterinary Medical Association. Chicago, 183:1302–1303.

Reid, R. D. 1936. Studies on bacterial pigmentation. Journalof Bacteriology 31:205–210.

Reid, J. D., Stoufer, S. D., Ogrydziak, D. M. 1982. Efficienttransformation of Serratia marcescens with pBR322plasmid DNA. Gene 17:107–112.

Reissbrodt, R., Rabsch, W. 1988. Further differentiation ofEnterobacteriaceae by means of siderophore-patternanalysis. Zbl. Bakt. Hyg. A 268:306–317.

Richard, C. 1972. Méthode rapide pour l’étude des réactionsde rouge de méthyle et Voges-Proskauer. Annales del’Institut Pasteur. 122:979–986.

Roussel, A., Lucas, A., Bouley, G. 1969. Serratia marcescens(Bacillus prodigiosus), une bactérie de la pathologiecomparée. Revue de Pathologie Comparée et de Méde-cine Expérimentale 6:27–29.

Sakazaki, R. 1974. Genus IX. Serratia Bizio 1823, 288, 326.Buchanan, R. E., and Gibbons, N. E. (ed.) In: Bergey’smanual of determinative bacteriology, 8th ed. Balti-more, Williams & Wilkins.

Salceda, S. R., Lapuz, J., Vizconde, R. 1973. Serratia marce-scens endophthalmitis. Arch. Ophthalmol. 89:163.

Schaberg, D. R., Alford, R. H., Anderson, R., Farmer III, J.J., Melly, M. A., Schaffner, W. 1976. An outbreak ofnosocomial infection due to multiply resistant Serratiamarcescens: Evidence of interhospital spread. Journal ofInfectious Diseases 134:181–188.

Schiebel, E., Braun, V. 1989. Integration of the Serratiamarcescens haemolysin into human erythrocyte mem-branes. Molecular Microbiology 3:445–453.

Sedlák, J., Dlabad, V., Motlikova, M. 1965. The taxonomy ofthe Serratia genus. Journal of Hygiene, Epidemiology,Microbiology, and Immunology 9:45–53.

Shinde, P. A., Lukezic, F. L. 1974. Characterization and sero-logical comparisons of bacteria of the genus Erwiniaassociated with discolored alfafa roots. Phytopathology64:871–876.

Sifuentes-Osornio, J., Ruiz-Palacios, G. M., Gröschel, D. H.1986. Analysis of epidemiologic markers of nosocomial


Serratia marcescens isolates with special reference to theGrimont biotyping system. Journal of Clinical Microbi-ology 23:230–234.

Sifuentes-Osornio, J., Gröschel, D. H. 1987. Modification ofGrimont biotyping system for epidemiologic studieswith nosocomial Serratia marcescens isolates. Journal ofClinical Microbiology 25:567–568.

Skerman, V. B. D., McGowan, V., Sneath, P. H. A. 1980.Approved lists of bacteria names. International Journalof Systematic Bacteriology 30:225–420.

Slatten, B. H., Larson, A. D. 1967. Mechanism of pathoge-nicity of Serratia marcescens I. Virulence for the adultboll weevil. Journal of Invertebrate Pathology 9:78–81.

Slotnick, I. J., Dougherty, M. 1972. Erythritol as a selectivesubstrate for the growth of Serratia marcescens. AppliedMicrobiology 24:292–293.

Smith, R. E., Hansen, H. N. 1931. Fruit spoilage of figs. Uni-versity of California Agricultural Station, Bulletin.506:1–84.

Smith, R. E., Reynolds, I. M. 1970. Serratia marcescens asso-ciated with bovine abortion. Journal of the AmericanVeterinary Medicine Association. 157:1200–1203.

Stanier, R. Y., Palleroni, N. J., Doudoroff, M. 1966. The aer-obic pseudomonads: A taxonomic study. Journal ofGeneral Microbiology 43:159–271.

Stapp, C. 1940. Bacterium rubidaeum nov. spec. Zentralbl.Bakteriol. Abt. 2 102:251–260.

Starr, M. P., Grimont, P. A. D., Grimont, F., Starr, P. B. 1976.Caprylate-thallous agar medium for selectively isolatingSerratia and its utility in the clinical laboratory. Journalof Clinical Microbiology 4:270–276.

Steigerwalt, A. G., Fanning, G. R., Fife-Asbury, M. A.,Brenner, D. J. 1976. DNA relatedness among species ofEnterobacter and Serratia. Canadian Journal of Microbi-ology 22:121–137.

Steinhaus, E. A. 1941. A study of the bacteria associated withthirty species of insects. Journal of Bacteriology 42:757–789.

Steinhaus, E. A. 1959. Serratia marcescens Bizio as an insectpathogen. Hilgardia 28:351–380.

Stephens, J. M. 1959. Immune responses of some insects tosome bacterial antigens. Canadian Journal of Microbiol-ogy 5:203–228.

Stephens, J. M. 1963. Bactericidal activity of hemolymph ofsome normal insects. Journal of Insect Pathology 5:61–65.

Stucki, G., Jackson, T. A. 1984. Isolation and characterisationof Serratia strains pathogenic for larvae of the NewZealand grass grub Costelytra zealandica. New ZealandJournal of Science 27:255–260.

Tagaki, T., Kisumi, M. 1985. Isolation of a versatile Serratiamarcescens mutant as a host and molecular cloning ofthe aspartase gene. Journal of Bacteriology 161:1–6.

Traub, W. H. 1972a. Bacteriocin typing of Serratia marcescensisolates of known serotype-group. Applied Microbiol-ogy 23:979–981.

Traub, W. H. 1972b. Continued surveillance of Serratiamarcescens infections by bacteriocin typing: Investiga-tion of two outbreaks of cross-infection in an intensivecare unit. Applied Microbiology 23:982–985.

Traub, W. H. 1980. Bacteriocin and phage typing of Serratia.79–100.von Graevenitz, A., and Rubin, S. J. (ed.) Thegenus Serratia. Boca Raton, CRC Press.

Traub, W. H. 1981. Serotyping of Serratia marcescens: confir-mation of five recently described new O antigens andcharacterization of an additional O antigen Zbl. Bakt.Hyg., I. Abt. Orig. A 250:307–311.

Traub, W. H. 1982. Opsonization requirements of Serratiamarcescens. Zbl. Bakt. Hyg., I. Abt. Orig. A 253:204–224.

Traub, W. H. 1983. Passive protection of NMRI mice againstSerratia marcescens: comparative efficacy of commercialhuman IgG immunoglobulin preparations and rabbitanti-O, -H, -K, -Life cell and -protease immune sera.Zbl. Bakt. Hyg., I. Abt. Orig. A 254:480–488.

Traub, W. H. 1985. Serotyping of Serratia marcescens: identi-fication of a new O-antigen (O24). Zbt. Bakt. Hyg., I.Abt. Orig. A 259:485–488.

Traub, W. H., Bauer, D. 1985. Degradation of humanfibronectin by metalloproteases of Serratia marcescens.Zbl. Bakt. Hyg. A 259:461–467.

Traub, W. H., Fukushima, P. I. 1979a. Further characterizationof “promptly” and “delayed” human serum-sensitivestrains of Serratia marcescens: simplified tube O-agglutination test and comparison with other serologi-cal procedures. Zbt. Bakt. Hyg., I. Abt. Orig. A 245:495–511.

Traub, W. H., Fukushima, P. I. 1979b. Serotyping of Serratiamarcescens: simplified tube O-agglutination test andcomparison with other serological procedures. Zbt.Bakt. Hyg., I. Abt. Orig. A 244:474–493.

Traub, W. H., Kleber, I. 1976. Selective activation of classicaland alternative pathways of human complement by“promptly serum-sensitive” and “delayed serum-sensitive” strains of Serratia marcescens. Infection andImmunity 13:1343–1346.

Traub, W. H., Kleber, I. 1977. Serotyping of Serratia marce-scens. Evaluation of Le Minor’s H-immobilization testand description of three new flagellar H antigens. Jour-nal of Clinical Microbiology 5:115–121.

Traub, W. H., Raymond, E. A., Startsman, T. S. 1971. Bacte-riocin (marcescin) typing of clinical isolates of Serratiamarcescens. Applied Microbiology 21:837–840.

Traub, W. H., Spohr, M., Bauer, D. 1983. Resistance-plasmid-and protease-independent murine virulence of a multi-ple-drug-resistant strain of Serratia marcescens. Zbl.Bakt. Hyg. A 256:184–195.

Trias, J., Viñas, M., Guinea, J., Loren, J. G. 1988. Induction ofyellow pigmentation in Serratia marcescens. Appliedand Environmental Microbiology 54:3138–3141.

Trought, T. E. T., Jackson, T. A., French, R. A. 1982. Theincidence and transmission of a disease of grass grub(Costelytra zealandica) in Canterbury. N. Z. J. Exp.Agric. 10:79–82.

Véron, M. 1975. Nutrition et taxonomie des Enterobacteri-aceae et bactéries voisines. Annales de Microbiologie126A:267–274.

Virca, G. D., Lyerly, D., Kreger, A., Travis, J. 1982. Inactiva-tion of human plasma "-proteinase inhibitor by a met-alloproteinase from Serratia marcescens. BiochimicaBiophysica Acta 704:267–271.

von Graevenitz, A. 1977. The role of opportunistic bacteriain human disease. Annual Review of Microbiology31:447–471.

von Graevenitz, A. 1980. Infection and colonization withSerratia. 167–186.von Graevenitz, A., and Rubin, S. J.(ed.) The genus Serratia. Boca Raton, CRC Press.

Ward, P. A., Chapitis, J., Conroy, M. C., Lepow, I. H. 1973.Generation by bacterial proteinases of leukotactic fac-tors from human serum, and human C3 and C5. Journalof Immunology 110:1003–1009.

Wasserman, H. H., Keggi, J. J., McKeon, J. E. 1961. Serrata-molide, a metabolic product of Serratia. Journal of theAmerican Chemical Society 83:4107–4108.


White, G. F. 1923a. Hornworm septicemia. Journal of Agri-cultural Research 26:447–486.

White, G. F. 1923b. Cutworm septicemia. Journal of Agricul-tural Research 26:487–496.

Wijewanta, E. A., Fernando, M. 1970. Infection in goatsowing to Serratia marcescens. Veterinary Record 87:282–284.

Williams, R. P., Hearn, W. R. 1967. Prodigiosin. 410–449.Gottlieb, D. and Shaw, P. D. (ed.) Antibiotics, vol. 2.Springer-Verlag. Berlin,

Williams, R. P., Qadri, S. M. H. 1980. The pigment of Serratia.31–79.von Graevenitz, A., and Rubin, S. J. (ed.) Thegenus Serratia. Boca Raton, CRC Press.

Wilson, C. D. 1963. The microbiology of bovine mastitis inGreat Britain. Bulletin de l’Office International des Epi-zooties 60:533–551.

Wright, C., Kominos, S. D., Yee, R. B. 1976. Enterobacteri-aceae and Pseudomonas aeruginosa recovered from veg-etable salads. Applied and Environmental Microbiology31:453–454.

Yamamoto, T., Ariyoshi, A., Amako, K. 1985. Fimbria-mediated adherance of Serratia marcescens strain US5to human urinary bladder surface. Microbiol. Immunol.29:677–681.

Yu, V. L. 1979. Serratia marcescens. New Engl. J. Med.300:887–893.

Zimmermann, L., Angerer, A., Braun, V. 1989. Mechanisti-cally novel iron (III) transport system in Serratia marce-scens. Journal of Bacteriology 171:238–243.

ZoBell, C. E., Upham, H. C. 1944. A list of marine bacteriaincluding descriptions of sixty new species. Bulletin ofthe Scripps Institution of Oceanography 5:239–281.

Documents

eTh unGe a The Genu Serratia - Semantic Scholar · et al., 1980). The type strains of “Bacillus indicus” (Eisen-berg, 1886), “Erythrobacillus pyosepticus” (Fortineau, 1904),