ON THE DECOMPOSITION OF AGAR-AGAR BY ANAEROBIC BACTERIUM'

SELMAN A. WAKSMAN AND W. BAVENDANINI

New Jersey Agricultural Experiment Station, New Brunswick, New Jersey, andInternational Expedition to the Bahamas

Received for publication, February 5, 1931

A study of the processes of decomposition of agar-agar bymicro6rganisms involves first of all a knowledge of the chemicalnature of the agar. This material is prepared from certainmarine algae and, due to the fact that it is used as a food in theOrient and for medicinal purposes in Europe and America, as wellas due to its employment for the preparation of culture media,its chemical nature has received considerable attention.

It has been found that the composition of the agar varies withthe nature of the plant from which it is obtained. Japanese agaris usually prepared from Gelidium corneum, while other species ofGelidium and species of Gracillaria, Fucus and Eucheuma are alsoemployed in different parts of the world.Although agar contains a considerable amount of ash, themajor

part of it consists of one or more hemicelluloses, or carbohydrateswhich are readily hydrolized by hot dilute mineral acids. Themost important constituent of the agar is a galactan, first referredto by Payen as gelose, to which he gave the formula C6Hjo05.Czapek (1913) in summarizing the results of previous investiga-tors, concludes that one-third of the agar consists of galactan. Ontreatment of agar with nitric acid, mucic and oxalic acids areproduced. Pentosans are usually reported to be present in agarbut in very small amounts. Fellers (1916) reported the presence

1 Journal Series paper of the New Jersey Agricultural Experiment Station,Department of Soil Chemistry and Bacteriology. This paper also forms con-tribution No. 6 of the International Expedition to the Bahamas.

91

JOURNAL OF BACTERIOLOGY, VOL. XXII, NO. 2

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

92 SELMAAN A. WAKSAIAN AND W. BAVENDAMM

of 3.12 per cent pentosan and 22.87 per cent galactan in purified(I3acto) agar. The galactan in the agar is usually referred to as6-galactan.The inorganic constituents are made up chiefly of Ca and Mg

sulfates. Without going into a detailed discussion of the v-olum-inous literature on the chemical composition and uses of agar(Fellers, 1916), a surnmary is given in table 1 of a few recentanalyses of this preparation. Attention is directed to the factthat the composition of the agar varies, depending upon the plant

TABLE 1

Chemical composition of agarPer cent of air-dry material

Moisture........................................... 15.29* 16.57tAsh ............................................ 4.23 3.85S........................................... 1.77 2.65Ca............................................. 0.66 0.92Mg............................................ 0.48 0.57Na............................................ 0.11 0.25K............................................. 0.11 0.07Ci............................................ 0.03 0.22p........................................... 0.02 0.05N............................................. 0.30 0.37Fat........................................... 0.37 0.30Fiber........................................... 0.89 0.80Carbohydrates (N-free extract) ...................... 77.34 76.15

* Forbes et al. (1913).t Fellers (1916). Average of 15 samples; ash constituents reported as oxides.

from which it was obtained, and upon methods of preparationand purification.Haas (1921) has shown that the ash content of carrageen ob-

tained from Chondrus crispus cannot be reduced by dialysisbelow 14.6 per cent; the ash consists principally of CaSO4. Insolutions of original carrageen, Ca-ions are present, but not sul-fate ions. Haas concluded, therefore, that the calcium and sul-fate ions are integral parts of the carrageen molecule and arepresent as a sulfuric ester:

O * So2 0

R Ca

0 * S02 * 0

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

DECOMPOSITION OF AGAR-AGAR

Neuberg and Ohle (1921) found that the sulfur present in theash of agar-agar was less than one-half of the total sulfur in theoriginal agar. Samec and Ssajevic (1922) suggested that theformula for agar was (C6HloO5)54SO4H. Fairbrother and Martin(1923) concluded that agar-agar consists principally of the calciumsalt of an acid sulfuric ester, namely (R 0 SO2 O)2Ca.

Several bacteria capable of liquefying agar have been isolatedin the past. Here belong the Bac. gelaticus of Gran (1902), theBact. betae-viscosum of Panek (1905), the Bac. Nenckii of Biernacki(1911), Microspira agar-liquefaciens of Gray and Chalmers (1924),Vibrio Andoi of Aoi (1925), Aoi and Orikura (1928), and others.Only one of these bacteria was isolated frorr sea water. Theothers were isoated from soil and other substrates. The specificbacterium is very abundant in sea waters, but has been leaststudied.Gran found that Bac. gelaticus produced an enzyme which

could hydrolize agar to sugar. Vibrio agar-liquefaciens decom-posed cellulose in addition to agar. The organism isolated byAoi and Orikura (1928) readily decomposed xylan, mannan andstarch, but did not decompose cellulose. This organism wasisolated by Aoi (1925) from manure. The bacterium did notdevelop on nutrient agar, but grew well on an inorganic mediumcontaining agar as a source of energy; it made good growth onGran's medium, whert 0.1 per cent NaCl was substituted for 3per cent; it also grew well with mannan as a source of energy.The organism was an obligate aerobe and was killed on heatingat 45°C. for ten minutes.H. and E. Pringsheim (1910) employed agar-agar as a source of

energy for nitrogen-fixation. They found that in the decomposi-tion of agar by Bac. gelaticus, there are formed substances whichreduce Fehling's solution and which are used by the nitrogenfixing bacteria as sources of energy. However, Lundestad(1938), who recently isolated an agar-destroying organism fromsea water, could not demonstrate any reducing sugars in theprocess of agar liquefaction. The quantitative destructionof the agar has so far not been studied.

93

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SELMAN A. WAKSMAN AND W. BAVENDAMM

EXPERIMENTAL

Occurrence and nature of organism used

In the course of a study of the microflora of marine sediments,an organism capable of energetic decomposition of agar, as indi-cated by the extent of liquefaction of an agar medium, was iso-lated. Hundreds of thousands of cells of this organism Nwerefound per gram of the mud sediment of William;s Island (BahamaIslands), especially in the upper regions of the mangrove-swamps,but also in the calciumn carbonate sedimenit around the coast. Theabundance of the organism could be very readily determined bymerely counting the number of depressions on the agar plateproduced by the specific organisms.The organisms developed best on the lactate-agar of Drew,

(1915), this medium having the following composition:gramns

Sea water..................................................... 1,000KNO3..............................................................0.5Na2HPO4................................................... 0.25Calcium lactate................................................... 2.00Agar.................................................... 18.00

The organism was isolated on a medium containing 0.5 gramK2HPO4 and 12.5 grams of agar in 1000 cc. of sea water. It wasbest cultivated upon a medium containing 0.2 gram peptone and10 grams agar in 1000 cc. of sea water. It grew, however, wellupon sea water agar with KNO3, (NH4)2SO4 or asparagin assources of nitrogen.

Colonies. Almost colorless and transparent, with yellowish-white surface. The bacterial slime is yellowish.

Morphology of organism. Small, motile rod, non-spore form-ing, 2 to 4 by 0.5 to O.7,u. Stains well with bacterial dyes.A bundance. One gram of moist mud containing 50,000 to

200,000 cells, as determined by plate method.The organism seems to be closely related to the Bac. gelatinus

isolated by Gran (1902), possibly being identical with it.22 A detailed study of the morphology of this organism and its relation to the

other agar-liquefying bacteria will be p)ublished later by the Junior author in theCentralblatt fur Bakteriologie (II Abteilung).

94

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

DECOMPOSITION OF AGAR-AGAR

Decomposition of agarFor the followinlg experiments the ordinary Difco Bacto Agar

has been employed. This agar had the following composition:per cent

AMoisture..................................................... 15.6Reducing sugar, on hydrolysis with 2 per cent HCI in flowingsteam for 5 hours............................................. 64.8

Ash..................................................... 2.3Nitrogen..................................................... 0.14

In the first experiment, 200 cc. portions of a medium containing1.0 per cent agar in sea water plus NaNO3 as a source of nitrogen(2 grams per liter) were placed in a series of 500 cc. Erlenmeyerflasks, sterilized, inoculated with 2 strains of the organism, andconnected with the respiration apparatus, for absorbing the CO2given off in the process of decomposition. Table 2 shows theamounts of C02 given off, measured as carbon. Table 3 gives thechemical composition of the cultures as compared with the con-trol, and brings out further the amount and nature of decomposi-tion that had taken place.The results presented in tables 2 and 3 show that considerable

decomposition of the agar has taken place in both cultures.Nearly a third to a fourth of the agar was liquefied in 31 days.Culture 2 was somewhat more active than culture 1. If theanalysis of the liquefied portion alone were presented, considera-bly greater decomposition would have been brought out; probablyover 80 per cent of the hemicellulose in the agar would have beenfound to have disappeared. However, since the amount of liquidavailable for analytical work was very small, and since the un-liquefied agar has probably also undergone partial decompositionand considerable diffusion of the salts has no doubt taken place,the total culture was hydrolyzed with 2 per cent hydrochloricacid in flowing steam and submitted to analysis.The two cultures brought about the decomposition of from 26

to over 37 per cent of the agar, as shown by the amount of hemi-cellulose that disappeared. On comparing the amount of hemi-cellulose decomposed (calculated as carbon content of the sugarround after hydrolysis) to the amount of carbon liberated as

95

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

96 SELMAN A. WAKSMAN AND W. BAVENDAMM

carbon dioxide, it is found that 71 per cent of the carbon of thehemicellulose decomposed in the agar has been liberated as gas,while 29 per cent has either been left in the fornm of intermediaryproducts or utilized by the organism for the synthesis of its cellsubstance. The fact that considerable synthesis has taken placeis brought out by the consumption of the nitrate-nitrogen andits transformation into organic nitrogen.

TABLE 2

Evolution of C02 in the course of decomposition of agar by a bacteriumMilligrams of carbon from 200 cc. of culture

DAYS OF INCUBATION'

CULTUItE TOTAL

8 10 12 15 19 22 25 31

Control .................... 3.2 2.7 5.9No. 1.14.8 17.8 14.3 22.2 22.6 16.6 108.7No. 2.................... 19.3 17.0 11.1 12.6 17.2 26.9 16.3 27.9 148.3

TABLE 3

Nature of agar decomposition by a bacterium

REDUCING NO N HEMICELLU- HEMI-SUGAR ON N AR

LOSE CARBON CELLULOSE C LIBERATEDHYDROLYSIS CONSUNMED* (AS SUGAR) AS CO2

CULTURE CO2 (AS CAR- CONSUMED NIT.ROGENLeft D)ecom- Left Con- DON) LITIER- NITROGENCONSSUMED

posed sumed ATED CONSUMED

?nfglln. ltZln. 7119711. mgm.

Control ....... 1.307 0 60.5 0No. 1......... 964 343 38.9 21.6 1.39 15.9 5.03No. 2......... 823 484 36.8 23.7 1.43 20.4 6.26

* Assuming that the carbon content of the hemicellulose is 44 per cent.

The fairly narrow ratio between the amount of hemicellulose inthe agar decomposed and the nitrogen assimilated points furtherto the abundant synthesis of bacterial cells.

In order to throw further light upon the metabolism of thisspecific organism, it was grown in a synthetic solution withammonium sulfate as a source of nitrogen and with glucose,starch and mannan as sources of energy. The media were steri-lized, inoculated and incubated at 270C. for 24 days. There wasvery little growth in all of these cultures; whether this was due

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

DECOMPOSITION OF AGAR-AGAR

to the low salt concentration (plain distilled water being used inthis experiment), or to the fact that ammonium salt was used as asource of nitrogen (this substance was found to be inferior tonitrate nitrogen for this organism), or to the inability of theorganism to attack these sources of energy, still remains to bedetermined. The data presented in table 4 point, however, tosome very interesting conclusions.The results show that only about 10 per cent of the carbohy-

drates were consumed by the organism. However, large quanti-ties of reducing sugars were produced from the starch and man-nan. This brings out the fact that even when conditions arenot very favorable for the growth of the organism, the latter

TABLE 4

Decomposition of different carbohydrates by an agar-liquefying bacterium

TOTAL SUGAR ONAMOURNT OF HYDROLYSIS WITH 2 PER

SUG}AR PRESENT CENT HCI SOLUTION

NATURE OF CARBOHYDRATES

Control Inocu- Inocu- Sugarlated lated sumed

mgm. mgm. mg,n. mgm. mgm.

Glucose.............................. 252 235 235 219 16Starch............................... 0 111 212 186 26Mannan ................................ 0 89.4 206 192 14

continues to produce an enzyme which is capable of hydrolyzingthe polysaccharides to reducing sugar and, as a result of the factthat the organism itself is not able to consume this sugar imme-diately, the latter continues to accumulate. This may be thereason why it is so difficult to demonstrate the production ofsugar in the process of cellulose decomposition by fungi andbacteria. When conditions are made unfavorable for the growthof the organism but favorable to enzyme action, the sugars willaccumulate (Payen, 1859, Pringsheim, 1912).

In order to obtain further information concerning the rateof agar decomposition by the agar-liquefying organism, especiallyas influenced by the nitrogen source, the results of anotherexperiment will be reported here (table 5). The medium con-

97

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

98 SELMAN A. WAKSMAN AND W. BAVENDAMM

sisted of sea water containing 1 per cent agar, 0.1 per cent K2HPO4and 0.2 per cent of the nitrogen source. Two hundred cubiccentimeter portions of the medium were placed in 500 cc. flasks;these were sterilized, inoculated with one strain of the organism,connected with the aeration apparatus and incubated for 90 days.The results show that the extent of decompositon of agar by

the agar-liquefying organism depends to a large extent uponthe nitrogen source. In the absence of available nitrogen, verylittle agar was decomposed, as indicated by the lack of liquefac-

TABLE 5

Influence of nitrogen source upon the amount and rate of decomposition of agar by anagar-liquefying bacterium

REDUCING C02 SOLUBLES UGAR O N GIVEN OFFiinI R GA NHYDROLYSIS NITROGEN

NITROGE(N SOURCE TREATMENT p

Cg ~-ga2)gC.mm.mContro1,334o

mgm. mQ,R. RliRl. mgmR. mgm. mgm.

Non Control 1,342 0 21.4 0 0 7.0

None.~~~~~Inoculated 1,291 51 43.3 21.9 0 0 7.0o(NH4)2SO4 ........

Control 1,334 0 21.6 79.2 7.0Inoculated 990 344 160.1 138.5 72.8 6.4 6.4

f Control 1,348 0 20.4 0 58.4 7.0NaNO3...Inoculated 602 746 214.0 194.0 34.1 24.3 6.9

tion, only a trace of CO2 being given off above the control, andby the limited reduction of the hemicellulose content in the agar.This fact proves that the organism is unable to use the combinednitrogen of the agar, as well as that it is unable to fix any atmos-pheiric nitrogen. The ammonium salt was found to be an inferiorsource of nitrogen for the bacteria as compared with nitrate.Although a considerable amount of agar was decomposed even inthe presence of the ammonium salt, the amount of cell substancesynthesized was considerably less, as shown both by the limitedamount of nitrogen consumed and the larger amount of carbon

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

DECOMPOSITION OF AGAR-AGAR

liberated as C02, when compared with a similar amount of de-composition in a much shorter period of time by culture I in theresults reported in table 3.The nitrate proves to be the best source of nitrogen for the

organism. More than twice as much carbohydrate is decom-posed, as shown by the reduction in hemicellulose content. Oncomparing these results with those reported in table 3, it is foundthat in this experiment there was more carbohydrate decomposedfor practically the same amount of nitrate-nitrogen assimilated.This is no doubt due to the longer incubation period, whichprobably leads to the autolysis of some of the cells with theresult that some of the nitrogen liberated is again assimilated bythe organism.

DISCUSSION

Agar-agar can be considered as a hemicellulose complex, in-cluding under the term "hemicellulose" those carbohydrateswhich are readily hydrolized by dilute acids, giving reducingsugars. Recent investigations have brought out the fact thatmost hemicelluloses so far known are not pure polysaccharides,but consist of mixtures of hexosans, pentosans and uronic acids.Agar-agar is such a hemicellulose, consisting largely of galactan,with an admixture of some pentosan and some uronic acid, aswell as certain inorganic salts. It has been shown elsewhere(Waksman and Diehm, 1931) that not all hemicelluloses aredecomposed alike by microorganisms. Some, like the mannansand certain pentosans, are very readily decomposed by a largenumber of microorganisms, while others, like the galactans,are very resistant to decomposition. Some of the hemicellulosesare more resistant to decomposition by microorganisms than thetrue cellulose. Agar-agar consists largely of a resistant hemicel-lulose, which is decomposed only by a few specific microorganisms,largely certain bacteria and actinomyces.A study of the decomposition of agar by a bacterium which

occurs in great abundance in sea mud, especially along the coastof various islands in the Bahamas and elsewhere, revealed thefact that the hemicellulose in agar can be readily utilized as a

99

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

SELMAN A. WAKSMAN AND W. BAVENDAMM

source of energy. In the process of agar decomposition, thebacterium synthesizes large quantities of microbial cell substanceand consumes large quantities of nitrogen. The bacterium canalso act upon other polysaccharides, such as starches and mlan-nans, and produce large quantities of sugar from these. Thisfact is of considerable interest, since it has a bearing upon itspossible symbiosis with other organisms, which can use this sugaras a source of energy, such as nitrogen-fixing bacteria, thus sup-plying both the agar-liquefying organisms and possibly the higherplants with available nitrogen.The growth of sea-weeds in a medium so poor in available

nitrogen, as the sea, thus becomes possible. This enables us topropose the following hypothesis:The algae groNing abundantly synthesize large quantities of

hemicelluloses.

The agar-liquefying and other similar bacteria decompose thesehemicelluloses liberating large quantities of available energy.

The nitrogen-fixing bacteria use this energy and fix atmosphericnitrogen.

The algae as well as the agar-liquefying bacteria use this nitro-gen, directly, or after the bacterial cells have autolized or under-gone decomposition.

SUMAMARY

A bacterium capable of liquefying agar was isolated from marinesediments around Williams Island. This bacterium was found tooccur in large numbers in the sediment.The agar-liquefying bacterium was found to attack rapidly

the hemicellulose complex of the agar and use it as a source ofenergy. A large part of the carbon was liberated as carbondioxide and a part was utilized by the organism for the synthesisof bacterial cell substance.

For the decomposition of the agar, the bacterium needs a

100

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

DECOMPOSTION OF AGAR-AGAR 101

source of nitrogen for the synthesis of its cell substance. Thenitrogen of the agar is not utilized readily. Nitrate nitrogen isa much better source of nitrogen than ammonium salts.The bacterium is capable of producing an enzyme which

hydrolizes mannan and starch to reducing sugars. Under un-favorable conditions of growth, the organism allows the sugarto accumulate.A theory is suggested to explain the role of the agar liquefying

bacterium in the cycle of life in the sea.

The organism was isolated from the Marine mud collectedduring the Expedition to the West Indies, led by Dr. Field ofPrinceton University. The authors are indebted to Dr. Field forthe facilities offered and for the participation in this Expedition.The authors are also indebted to Mr. Reuszer and Mr. Purvis ofthis laboratory for assistance in making the chemical analyses.

REFERENCESAoi, K. 1925. tber eine neue agarzersetzende Bodenbakterienart. Vorlaufige

Mlitteilung. Centrbl. Bakt. II, 63, 30-32.Aoi, K., AND ORIKURA, J. 1928. On the decomposition of agar, xylan, etc. and

the sugars related to these hemicelluloses by Vibrio andoi (n. sp.).Centrbl. Bakt. II, 74, 321-333.

BIERNACKI, W. 1911. Bacterium Nenckii Biern., ein neuer den Agar verflus-sigender Mikro6rganismus. Vorlaufige MIitteilung. Centrbl. Bakt.II, 29, 166-169.

CZAPEK, F. 1913. Biochemie der Pflanzen, 1, 643.DREW, G. H. 1915. On the precipitation of the calcium carbonate in the sea by

marine bacteria, and on the action of denitrifying bacteria in tropicaland temperate seas. Carnegie Institute, Washington, 5, 7-45.

FAIRBROTHER, F., AND M\IARTIN, H. 1923. The swelling of agar-agar. Jour.Chem. Soc., 123, 1412-1424.

FELLERS, C. R. 1916. The analysis, purification and some chemical propertiesof agar-agar. Jour. Ind. Engin. Chem., 8, 1128-1133.

FORBES, E. B., BEAGLE, F. MI., AND MIENSCHING, J. E. 1913. iMineral andorganic analyses of foods. Ohio Agr. Exp. Sta. Bull. 255.

GRAN, H. H. 1902. Studien uiber Meeresbakterien. II. Iber die Hydrolysedes Agar-Agars durch ein neues Enzym, die Gelase. Bergens MuseumAarbog, No. 2, 1-16.

GRAY, P. H. H., AND CHALMERS, C. H. 1924. On the stimulating action ofcertain organic compounds on cellulose decomposition by means of anew aerobic microorganism that attacks both cellulose and agar. Ann.Appl. Biol., 11, 324-338.

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

102 SELMiAN A. WAKSMAN AND W. BAVENDAMAM

HAAS, P. 1921. Carragen (Chondrus crispus). II. The occurrence of etherealsulfates in the plant. Biochem. Jour., 157, 4694-76.

LUNDESTAD, J. 1928. lUber einige an der norwegischen KiIste isolierte Agar-spaltende Arten von AIeerbakterien. Centrbl. Bakt. II, 75, 321-344.

NEUBERG, C., AND OHLE, H. 1921. tber einen Schwefelgehalt des Agars.Biochem. Ztschr., 125, 311-313.

PANEK, MI. K. 1905. Bakteriologische und chemische Studien ilber die "Bar-szcz" genannte Giarung der toten Ruiben. Bull. Acad. Sci. Cracovie, 1,5-49.

PAYEN, C. R. 1859. (Reference by Fellers.)PRINGSHEIM, H. 1912. tiber den fermentativen Abbau der Zellulose. Ztschr.

physiol. Chem., 78, 266-291.PRINGSHEIM, H. 1912. tber den fermentativen Abbau der Hemicellulosen. I.

Ein Trisaccharid als Zwischenproduct der Hydrolyse eines Mannans.Ztschr. physiol. Chem., 80, 376-382.

PRINGSHEIM, H., AND E. 1910. lber die V-erwendung von Agar-Agar als Ener-giequelle zur Assimilation des Luftstickstoffs. Centrbl. Baktl. II,26, 227-231.

SAMEC, MI., AND SSAJEVIC, V. 1922. Noll. Chem. Beihefte, 16, 285-300.VAN DER LEK, G. B. 1929. Vibrio agarliquefacians Gray. Ned. Tijdschr. Hyg.,

MIicrobiol., Serol., 3, 276-280. (Ref. Centrbl. Bakt. JJ, 79, 435-436.)WAKSMAN, S. A.., AND DIEHM, R. A. 1931. On the decomposition of hemicellu-

loses by micro6rganisms. Soil Sci., (to be published).

on Decem

ber 19, 2020 by guesthttp://jb.asm

.org/D

ownloaded from