lHIlllHlllllllllll|||l|lllllllllllllllllllllllllllllllllllll

3 1293 10524 7922

This is to certify that the

thesis entitled

Uric Acid and Uricolytic Gut
Bacteria in Termites:

A Strategy for Nitrogen Conservation
presented by

Catherine J. Potrikus
has been accepted towards fulfillment

of the requirements for

a Ph.D. degreein Microbiology

Date - 2 /

0-7639

 

 

 
    

hw- ' _.

. mutt“

   
  

% was
~ LIBRARY 1
M'C‘firn“ 3:258 {g

.‘ Uhi‘t’r‘)‘ “F
JJLL “I'M

OVERDUE FINES:
25¢ per day per item

lf""‘\\\ t RETURNING LIBRARY MTERIALS:

Place in book return to reuove
charge from circulation records

“94f

 

at. f 8 3.5:; ‘ g

 

 

URIC ACID AND URIGDLYTIC GUT BACTERIA
IN WOOD-EATING TERMITES:

A 9mm FOR NITROGENCINSERVATIQR

By

Catherine Jane Potrikus

A DISSERTNTION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCIOR.OF PHILOSOPHY

Department of Microbiology and Public Health

1980

v], V

('5’ / / U"

URIC ACID AND URICDLSTTC GUT BACTERIA
IN WUODbEATING TERMITES:
.A STRATEGY FOR.NITRDGEN CONSERVATION

By

Catherine Jane Potrikus

Uric acid (UA) was present in termites as assessed by paper and
thin layer chromatography, by U.V. spectroscopy, by reactivity with
uricase, and by gas chromatography coupled to mass spectranetry.
Specimens of six species of termites (Reticulitermes flavipes, 3.
virflnicus, Ooptotermes formosanus, Marg initermes hubbardi ,
Paraneotermes sinplicicornis, and Cryptotermes cavifrons) contained DA
in amounts accounting for between 1 and 45% of the termites’ dry weight
and for between 4 and 69% of the termites’ total nitrogen. Almst all
of the UA in 3. flavigs was associated with fat body tissue. During
15 months of laboratory captivity, UA in 5. flaviEs workers increased
fran 1.7 to 45.5% of the insects’ dry weight, apparently at the expense
of some endogenous nitrogen and carbon. attracts of g. flavigs
abdominal tissues possessed purine nucleoside phosphorylase and
xanthine dehydrogenase activities to synthesize UA, but 5. flaviEs
tissues lacked uricase as well as non-ur icase mediated pathways to
degrade UA. However. little or no UA was voided in termite feces.

Uricolytic bacteria were present in guts of g. flaviEs in populations

Catherine Jane Potrikus

of up to 6 x 104 cells/gut. Of 82 strains isolated under strict

anaerobic conditions, most were group N Streptococcus sp., Bacteroides

 

 

termitidis, and Citrobacter sp. All isolates used UA as an energy

 

source anaerobically. but not aerobically. Streptococcus strain UAD—l

 

degraded UA inconpletely unless formate (or a formicogenic coupound)

2(Eo g

-430mV) also provided 28+ + 2e- which served to drive ur icolysis.

was present as a co-substrate. Formate oxidation to CD

Streptococcus UAD-l degraded UA as follows: lUA + l formate -—--->

«Dz + l acetate + 4NH3. Formate dehydrogenase and uricolytic

activities of strain UAD-l were inducible by growth on UA. B.

termitidis UAD-SO degraded UA as follows: lUA > 3.50) + 0.75

2
acetate + 4NH3. Uricolytic activity was inducible in strain UAD—SO.

 

 

However, exogenous formate was neither required for nor stinulatory to

uricolysis by this strain. Studies of [IA catabolism by Citrobacter

 

strains were limited, because only small mounts of HA were consumed by

the cells in liquid culture or by resting cell suspensions.

Streptococcus UAD-l and _B_. termitidis HAD-SO readily evolved 14002

 

from [2-140] UA under anaerobic conditions. Thus, [2-14C] UA was

used as a substrate to measure uricolysis by termite gut preparations
and by whole termites. Fresh termite gut contents degraded UA
anaerobically, but not aerobically, at rates of up to 82 pmles x

1 1

gut- x h- . No UA was degraded by washed segments of gut tissue.

Termite Malpighian tubules transported UA it; £3.39! and when [2-14C]
UA was injected into 3. flaviEs workers “(Dz was evolved,
indicating in v_i§ transport of HA from the hemolynph to the gut for
anaerobic catabolism by the resident microbiota. Products of UA

catabolism by gut bacteria may be useable by the termite for nitrogen,

Catherine Jane Potr ikus

carbon, and energy. Crude extracts of 5. flaviEs tissue possessed
glutamine synthetase activity, suggesting that termites are able to
assimilate N83. the major nitrogenous product of gut microbial

uricolysis. Recycling of HA nitrogen by gut bacteria thus appears to
be an important means of nitrogen conservation for termites: the
ammt of nitrogen liberated by gut uricolysis could permit a
population increase of at least 20% annually. Although generally
regarded as an insect ”waste product", uric acid may be an inportant
reserve material in termites, useable only with the aid of their gut

microbiota.

© Copyright by
CATHERINE JANE MRIKUS
1980

'Ib my Mother, and in memory of my Father,
for their love, their support, and their trust in me,
even when they did not understand.

iii

I wish to thank the members of my guidance cannittee, Dr. J. M.
Tiedje, Dr. C. A. Raddy, Dr. M. J. King, and Dr. D. G. Cochran
for their advice and assistance during the course of these studies. My
special thanks to Dr. Cochran for his ready replies to letters and
phone calls for advice on insect physiology.

I express deep appreciation to some very special friends and
colleagues, those with whom I shared the peaks of elation and pits of
despair that accarpany scientific research: Larry Forney, who kept me
sane through countless uric acid assays, and was always a willing (P)
listener, a ready critic (in the most favorable sense of the word), and
a very special friend; Jon Wegienek, whose friendship likewise has
been a source of strength and encouragement; and Bob and Ellen Uffen
et a1, without whan I never would have quite so happily retained my
contact with family life - thank you for Sunday dinners, a garden
plot, kids and pets, and most of all for your sincere friendship,
advice, and emcuragement.

Finally, and rust sincerely, I wish to thank Dr. John A.
Breznak, my major professor, for the many, many hours of everything
that goes into holding a scientist. thank you for your patience, your
encouragement, your advice and guidance, and particularly for sharing

with me your great enthusiasm for scientific research.

iv

TABLE CF WIS
page
LI” m mm 0 O O O O O O O O O O O O O O O O O O O O O O O O O 0 Vi

LISTCFFIGURESvni
Insect-MicrobeSynbioses..................4

mles of microbial synbionts in insect nutrition . . . .5

NitrogenEccruuyofTermites................16
STATWI'CFPURPCBE........................25
LITERATURECITH)..........................28
ARI‘ICLEI..36

Uric Acid in Wood-Eating Termites
ARI‘ICLEII46

Uric Acid-Degrading Bacteria in Guts of Termites
[mticulitermes flavipes (Kollarll

 

mxaE III 0 O O O O O O O O O I O O O O I O I O O O O O I I I O O .55
Anaerobic Degradation of Uric Acid by Gut Bacteria of Termites
MI“ IV C O I O O O O O O O I O O O I O O O O O O O O O O O O O O 64

Gut Bacteria Degrade Uric Acid: A Strategy for
Nitrogen Conservation in Termites

”MIX I O O O O O I O O O O O I O O O O 0 O O O O O I O O O O O O 100
Sunmary Statanent
MIX I I O O I O O O 0 O O O O O O O O O O O O I O O O O O O O O 10 5

Nitrogen-fixing Ehterobacter agglomerans Isolated from
Guts of Wood-Eating Termites

V

Table

ARI'ICLEI

ARTICIE II

ARTICLE III
1

ARI'ICLEIV

LIST (F RELES

page

Uric acid (DA) and total nitrogen (N) content
of 3; flavigs workers during captivity . . . . . . . . .39

Uric acid (DA) and total nitrogen (N) content
ofvarioustermitespecies. . . . . . . . . . . . . . .40

Uricolytic bacteria present in guts of g: flavipgs . . . 50

Characteristics of uricolytic group N Streptococcus
isolatesfrangtflavias. . . . . . . . . . . . . . . .Sl

 

Anaerobic utilization of uric acid and other hetero-
cycliccarpoundsbygrowingcells. . . . . . . . . . . .52

Anaerobic metabolism of [2-14C] UA by minced guts of
B:- flaViEs I I I I I I I I I I I I I I I I I I I I I I 053

Fermentation of m and fructose by growing cells of
Strwms W]- I I I I I I I I I I I I I I I I I I I58

Fermentation of UA and [II-MC] fructose by cell suspen-
sionsofStreptococcusUAD-l..............59

Fermentation of HA and [14C] formate by cell suspensions
ofStreptococcusUAD—l.................60

Fbrmate oxidation by cell suspensions of Streptococcus
DAD-1 I I I I I I I I I I I I I I I I I I I I I I I I I I61

Permutation of DA by cell suspensions of B. termitidis
DAD-50 I I I I I I I I I I I I I I I I I I I I I I I I I 61

 

14C—ur ic acid production fran 14C-xanthine by 3.

flaviEsabdanenextracts................'78

Absence of uricase in tissues of R_. flaviEs . . . . . . 80

vi

Table page

3 14(1)2 evolution fron [2—14C] uric acid by 3.
flaViEsgutpreparations. . . . . . . . . . . . . . . .82
APPHQDIX II
1 Biochemical reactions of strains C-1 and C-2 . . . . . .108
2 Ebrmentation products of termite isolates and

§_._aggloneransClr811-74...............108

 

3 Specific growth rates of g. agglonerans strains
isolatedfrontermites. . . . . .. . . . . . . . . . .108

4 Effect of anaerobiosis on C H reduction by
.C_._ formosanus and their extgagted guts . . . . . . . . .110

 

vii

Figure

AM‘ICLE I

ARTICLE III

1

ARTICLE IV

APPH‘IDIX I

APPENDIX I I

LIST (1“ FIGURES

page

U.V. absorption spectra of Li (I) extracts of

standard uric acid and of 5. flagigs workers

before and after exposure to oomercial hog
liveruricaseenzymefor30min. . . . . . . . . . . . .38

Mass spectrum of tetraethyl uric acid prepared
frmmrkersof&flavig...............38

Uric acid (DA) and total nitrogen (N) content
of §_._ flavies during laboratory captivity. . . . . . . .39

Deposition of uric acid in fat body tissue of
B_.f1aviEsworkers...................42

(1) evolution from fructose (F) and CA by cell
an nsimsofStreptococcusUAD—l. . . . . . . . . . . .58

 

Fructose-driven UA dissimilation by cell sus-
pensionsofStreptococcusUAD—l. . . . . . . . . . . . . 59

. Fbrmate-driven UA dissimilation by cell suspensions

ofStreptocoocustJAD—l..................60

Uptake of [2-14C] uric acid by Malpighian tubules
Of -R_I. flaVi-ESI I I I I I I I I I I I I I I I I I I I I I85

Evolution of 14(I) fron [2-14C] uric acid injected
intothehemol of&f1wiEs............87

A strategy for microbial-mediated conservation of
conbined nitrogen in termites. . . . . . . . . . . . . .102

Growth and C H reduction exhibited by E.
ggloneransgtgainC-Z.................109

02 inhibition of C Hz reduction by cells of
fimla'erans St: inc-10 o o o o o o o o o o o o a o 0109

viii

IWIQV

Termites are insects belonging to the order Isoptera. They are
social animals whose survival depends on a precisely defined division
of labor between various caste members of the colony. The worker caste
or functional group doninates the activities of a colony, in part
because of their large numbers. Workers are the primary builders and
foragers, and they supply nutrients to reproductives, soldiers, and
ymng (nutrient-dependent) larvae. The intricacies of termite survival
extend beyond cormunal interactions, however, and also involve
interactions with symbiotic microorganisms. In fact, one of the most
oft-cited examples of nutualism in biology is the interaction of
evolutionarily "lower" termites* with their cellulolytic hindgut
protozoa. The latter organisms enable xylophagous termites to digest
cellulose (the major polysaccharide of wood), and the phenonenon lies
at the very heart of the termites’ carbon and energy nutrition.

Although Cleveland (23,24) long ago recognized the termite’s
dependency o1 gut protozoan symbionts, it was Hungate (48,50) who
* The order Isoptera includes five living families: Mastotermitidae,
Kalotermitidae, Hodotermitidae, Rhinotermitidae, and Termitidae (103) .
The phylogenetically more primitive termites, mothers of the first four

families, are referred to as lower termites, whereas the Termitidae are
regarded as higher termites.

elucidated the role of protozoa in cellulose metabolism. Hungate
demonstrated that suspensions of mixed microorganisms (containing

primarily protozoa) fron Zootermcpsis fermented cellulose to acetate,

 

CO , and Hz. Acetate produced in the gut was absorbed throigh the

2

gut wall, and the termite oxidized the acid to CI) and H O for

2 2
energy (50). Hungate’s elegant early work has recently been verified

by Yamin (107,108) who isolated and maintained the protozoan

Tr ichomitopsis termops idis from Zootermops is . By using

14C—cellulose, Yamin (108) showed that pure cultures of T.

 

termcpsidis fermented cellulose to acetate, 002, and H2. However,

as Yamin (108) points oit, this protozoan is not the major cellulolytic
species in the termite hindgut. Although progress has been made in
studying the intriguing protozoan symbionts from termites, further
investigations are needed to fully elucidate their roles in termite
nutrition.

In addition to protozoa, an abundant and diverse bacterial
population resides in hindguts of lower termites (11). Unfortunately,
the procaryotic component of the gut microbiota has been largely
ignored. Investigators have only recently begun to elucidate some of
the roles termite gut bacteria have in the nutrition of their host
(ll,12,38,39,9l,92). One intriguing role, discussed further below,
concerns N2 fixation by termite gut bacteria. The ability of
termites to thrive on relatively N-poor diets proupted investigations
for bacteria capable of fixing atmospheric N2 to supplement the
insect’s meager N intake. Although these studies proved fruitful, for
NZ-fixing bacteria were found in termite hindguts (39,86), the

results did not always reconcile the discrepencies that existed between

termite dietary N intake and the presumed N requirements for termite
colmy maintenance and growth. It appeared that other means for
conserving N could be important to wood-eating termites. Thus, after a
preliminary investigation which resulted in the isolation and

identification of Nz-fixing Enterobacter agglomerans from guts of

 

Cgptotermes formosanus termites (see Appendix II), the present study

 

was undertaken to evaluate the role of gut microorganisms in recycling
termite nitrogenous excretory conpounds. Results presented herein show
that gut bacteria conserve termite N by degrading uric acid (a presumed
waste product which is synthesized by the termite) to products
utilizable by the host. This role of microbes in termite nutrition
adds yet another dimension to oar growing knowledge of an array of

intriguing termite-microbe symbioses.

LITERATURE REVIEW

Insect : Microbe Symbioses

synhiosis, as defined by deBary (5) , is an intimate and relatively
stable association of two dissimilar organisms. Buchner (18) and
Richards and Brooks (89) elaborated this definition in describing
endosymbiosis, wherein one organism is sheltered within the body of
another, with the arrangement mtually beneficial to both the host and
the symbiont. Symbioses between insects and microbes encompassed by
this latter definition embrace a myriad of morphological,
physiological, and biochemical interactions, including both
intracellular and extracellular microbial habitats within the insect
host. Morphological and behavioral aspects of numeron insect -
microbe symbioses have been well documented (13,18,58,105). By
contrast, the specific roles of microbes in their hosts’ nutrition,
growth, and reproductim have largely remained speculative, and
therefore the inportance of the symbionts to their hosts has not, in
most cases, been clearly elucidated.

Morphological adaptations between insects and their symbionts are
diverse (89). Microorganisms may be harbored in specialized cells
(mycetocytes) scattered throighout the insect’s fat body tissue; in

specific organs (mycetones) designed to harbor and/or transmit the
4

symbionts; in Malpighian tubules; in various specialized or
non-specialized portions of the digestive tube; or in the reproductive
organs. Insects which harbor microbial symbionts may be dependent upon
them for survival, or the insect may benefit fron the association, but
survive in its absence.

The physiological significance of insect endosymbionts can best be
studied by removal of symbionts from their host. The symbiont-free, or
aposymbiotic, insects are then conpared with symbiont-containing
counterparts, often under conditions of dietary or environmental
stresses. Ideally, the symbionts would also be isolated and cultured,
and their ifl 112:2 characteristics and physiology correlated with their
predicted roles E gi_v_o_ to affirm the microbes’ significance to their
host.

Koch (57) has discussed methods frequently used to induce
aposymbiosis in insects. Symbionts may be eliminated, or their
acquisition by young prevented, by: surface sterilization or physical
perturbation of eggs; surgical removal of symbiont-containing
mycetones; incubation of insect hosts at elevated temperatures;
treatment with the enzyme lysozyme; or injection or feeding of insects
with antibiotics (57) . The efficacy of such methods depends upon the
individual insect species under investigation and the characteristics
of the symbionts and of the symbiosis. Eacploitation of aposymbiotic
specimens has helped elucidate many intriguing roles of microbial

symbionts in insect nutrition (18,57,105).

mles 3f microbial gimbionts _i_n insect nutrition

 

Symbionts play significant roles in supplementing nutritionally

inadequate diets for their hosts (18,89,105). Richards and Brooks (89)

pointed th a frequent correlation between an insect’s food source and
the presence or absence of symbionts. For example, insects which
thrive on diets of wood, plant sap, or animal blood often harbor
syrrbionts (89). Such diets are deficient in N and/or vitamins. The
hypothesized roles for the symbionts are to assist in digestion of the
food, or to supplement diets with vitamins or corbined N (e.g., amino
acids) required for the growth and development of the insect host.
Digestion of wood or cellulose is frequently mediated by microbes
foind either external to the wood—eating insect (i.e., infesting the
wood itself) or harbored within the insect as endosymbionts (37,89).
As mentioned in the Introduction, some xylophagous termites rely on
cellulose fermentation by their associated gut protozoa for a source of

carbon and energy. When Reticulitermes flavijes termites were

 

defaunated (i.e., rid of their protozoa) by incubatim at high
tenperature or under hyperbaric oxygen, or by starvation, they could
not survive am a normal diet of sound wood (23,25). However,
refaunation, or feeding the termites fungus-degraded wood, prolonged
survival indefinitely, implicating the gut protozoa in wood digestion
in li_v_g. Moreover, Hungate (47) demonstrated that cellulase activity

in Zootermcpsis was confined to the hindgut of normally faunated

 

specimens, and was absent fron defaunated termites. These data, and
the recent isolation and study of cellulolytic protozoa fron termites
(107,108), confirm that the protozoa are inportant to some termites’
nutrition.

Symbiotic cellulolytic protozoa are not found in the hindguts of
all termites. In the higher termites, devoid of cellulolytic protozoa,

cellulase production has been ascribed to gut bacteria (61,71,84,98)

or, in the fungus-growing Macrotermes, to fungal conidiophores (1,68) .

 

Evidence for cellulase production by the higher termites themselves has
also been presented (1,68,87) , but usually in concert with bacterial or
fungal cellulolytic activity. It appears, therefore, that most
termites are assisted in their digestion of cellulose by symbiotic
associations.

Gut protozoa of the wood-eating cockroach Cryptocercus punctulatus

 

ferment cellulose to reducing sugars which are presumably utilizable by
the host (99) . Trager (99) and Cleveland (27) denonstrated that

cellulase activity in normally faunated ngptocercus was localized in

 

the hindgut, where the protozoa reside, whereas defaunated individuals
were devoid of cellulase activity. These results implicate the gut
protozoa of Cryptocercus in cellulose metabolism in a manner similar to
the role ascribed to termite protozoa.

In contrast to the endosymbiotic cellulolytic protozoa of termites

and gyptocercus, cellulolytic bacteria are important in the nutrition

 

of wood—eating Lamellicorn beetle larvae. The bacteria are ingested
and retained with the beetle’s food in a specialized sac in the hindgut
(104). After digesting the wood, the microbes are themselves consumed
by their host (104). Although this relationship is not a permanent
endosymbiosis as observed with termites, the role of bacteria in
digestion is critical to the beetles. Cellulolytic bacteria likewise
assist other species of xylophagois insects in their digestive
processes (see 18, 105).

Perhaps nore intriguing than their roles in digestion are the
roles that insect-associated microbes play in their hosts’ N nutrition.

Insects whose diets are deficient in vitamins and amino acids may

8

depend upon symbionts to provide these essential nutrients. Moreover,
insects whose diets are N deficient may rely on microbes for N
acquisition (e.g., N2 fixation) or for conservation of combined N.
Many terrestrial insects synthesize uric acid as a product of protein
and nucleic acid catabolism (19,22,29) . Uric acid is a non-toxic
nitrogenous ccnpound which can be eliminated with minimal water loss to
the insect, and thus it is an ideal excretory waste for insects which
must conserve water (19,22,29) . However, for insects living on N—poor
diets, the loss of N in any form could be detrimental. It would be
advantageous if such insects possessed means to conserve uric acid N
which might otherwise be lost during defecation. In fact, some insects
appear to have evolved strategies for conserving uric acid N. Even
cockroaches, which do not normally live on N-pcor diets, can accumulate
large amounts of uric acid reserves when dietary N is in excess. The
potential roles of microbial symbionts in insect N nutrition, and
problems in elucidating symbiont roles, are well illustrated by studies
with urate-storing cockroaches.

The fat body tissue of insects functions as a site for storage of
metabolites, including fats, carbohydrates, and proteins (22). In many
insects, this tissue is also the site of uric acid synthesis and
storage (29,67) . In cockroaches, the purine can account for as web as
30% of the insect’s dry weight and 72% of the total body N (31) .
However, whereas uric acid was once regarded solely as a waste product,
and its sequestration in fat body tissue termed "storage excretion"
(29,67), recent evidence suggests that the stored purine is in a
dynamic state in cockroaches (29) .

Urates may provide cockroaches with N reserves utilizable in times

9

of dietary stress. Haydek (43) first documented the effect of dietary
N on the uric acid content of cockroaches. men specimens of
cockroaches were fed high protein diets, they accumulated large masses
of white crystals, presumably uric acid (43). Cockroaches fed N—free
diets of dextrin apparently lost the white deposits (43). In fact,
Haydek surmised that insects fed diets containing 79% to 86% protein
were killed by the massive deposits which accumulated throughoit their
bodies and interferred with the function of other organs. Although
this conclusion may have been premature, pending further nutritional
studies (43,75) , cockroaches do appear to store uric acid when dietary
N is in excess.

Haydek’s findings have been verified and elaborated by Mullins and
Cochran. Most cockroaches do not excrete uric acid (28,30,3l,77) , even
under dietary stresses which cause uric acid concentrations to

fluctuate significantly (74). When Periplaneta americana were fed

 

diets with a N content greater than the insects’ N needs, excess N was
stored as uric acid (75) , sometimes even at the expense of carbon
reserves. Furthermore, insects fed low N diets nobilized their urate
stores (76). The rate of uric acid nobilization was related to the
level of dietary N, and was always greater in females than in males,
presumably because of the females’ higher N demand during oothecal
production (76). In fact, Mullins and Neil (78) have shown that uric

acid synthesized by the male German cockroach [Blattella germanica

 

(L.)] and voided from accessory glands mto spermatophores is
incorporated into female oothecae. The mount of
spermatophore-asscciated urates incorporated into the oothecae appears

to be related to the female’s dietary N, and thus to her need for an

10

additional exogenous N supply (78). Clearly, uric acid can be more
important to cockroach N nutrition than merely serving as a nitrogenous
waste.

The mechanism(s) of urate mobilization in cockroaches is not well
understood. meiotic bacteria, both free-living in the gut and
intracellular in fat body myoetccytes, have been implicated in the
mobilization and subsequent catabolism of uric acid (16,73) .
Symbiont-free cockroaches have been useful experimental animals for
examining the impact of microbes in cockroach N metabolism. The
progeny of cockroaches fed an antibiotic such as Aureomycin for three
to four months are aposymbiotic, i.e., freed of their mycetocyte
symbionts (15). Brooks and Richards (16) compared urate levels in
symbiont-free fat body and normal fat body of B. germanica after
transplanting the latter tissue into an aposymbiotic host. The normal
tissue, which remained healthy and which contained abundant
bacteria-filled mycetocytes, was relatively free of urates 127 days
after implant, whereas the symbiont-free tissue accumulated massive
mounts of urates during the same period (16) . There was no infection
of aposymbiotic fat body by bacteria from the implanted tissue. The
authors concluded that the accumulation of uric acid was inhibited, or
urates were utilized, in tissue containing mycetocyte bacteria (16).
The role of the bacteria could not be further defined.

Pierre (83) observed uric acid stores in aposymbiotic cockroaches

of the species Leuccphia maderae. Moreover, he reported uricase

 

activity (i.e., the aerobic conversion of uric acid to allantoin) in
suspensions of isolated mycetocyte symbionts (83) . Likewise, Donnellan
and Kilby (35) reported that bacteria isolated from the fat body of g.

11

americana degraded uric acid and allantoin aerobically to glyoxylate
and urea; the latter corpoand was subsequently catabolized to NH3

and (112. However, the work of Pierre, Donnellan and Kilby, and
others who have reportedly isolated and studied cockroach mycetocyte
symbioits (e.g., see 17) has been challenged by Brooks and Richards
(17). As these latter authors point out, the techniques for symbiont
isolation were so stringently designed to insure aseptic conditions
that the symbionts themselves were undoabtedly killed during the
isolation procedures. The cultured ”symbionts" were, therefore,
contaminants (17) . Moreorer, studies with aposymbiotic cockroaches
were not always valid. As mentioned above, progeny of antibiotic-fed
cockroaches were free of mycetocyte bacteria. The parental generation,
however, usually retained some symbionts, even during longterm (i.e.,
foJr months) antibiotic feeding (15,17). When these insects, rather
than their progeny, were used for experimentation, the results did not
reflect true aposymbiotic conditions (17). With these criticisms in
mind, it appears that the roles of symbionts in cockroach urate
metabolism must yet be rigorously investigated. Nevertheless, the
presence of bacteria—filled mycetocytes in close proximity to urate
cells in cockroach fat body (32), transovarian transmission of the
symbionts to oocytes (l4) , and the possibility that cccyte uric acid is
metabolized during embryogenesis (78) further support the contention
that mycetocyte bacteria may be involved in cockroach urate metabolism.
Although antibiotic-fed cockroaches may not be truly aposymbiotic,
their symbiont population is greatly reduced (15). Studies comparing
these insects with normal cockroaches have shown that the symbionts do

indeed affect their host’s nutrition. "Aposymbiotic" cockroaches,

12

whether parental antibiotic-fed specimens or their progeny, are much
lighter in color than their normal counterparts, grow more slowly and
molt less frequently, and produce few, if any, fertile eggs
(15,16,66,89) . Richards and Brooks (89) caution that these effects may
be a result of the antibiotics themselves. Significant, however, was
Brooks and Richards’ (16) report that implantation of normal fat body
into aposymbiotic _B_. germanica restored the insects’ normal coloring
and growth.

The effects of aposymbiosis on cockroaches have been ascribed to
the lack of essential vitamins and amino acids that are normally
supplied by the mycetocyte bacteria. Henry and Block (45) fed
[U-MC] glucose to normal and aposymbiotic B. germanica and analyzed
the production of 14C-labeled amino acids. Aposymbiotic insects
failed to synthesize eight amino acids that were present in normal
specimens (45). Brooks and Kringen (14) found that polypeptides
present in hydrolyzed lactalbumin, proteose peptone, brewer’s yeast,
and lyophilized mouse liver enabled symbiont-free B. germanica
specimens to grow at a more normal rate than specimens fed on a casein
stock diet. However, the authors could not further characterize the
growth factor, nor determine the role of symbionts in supplying it to
the host (14).

Ludwig and Gallagher (66) found that the fat body of aposymbiotic
_P. americana contained reduced mounts of ascorbic, folic, and
pantothenic acids when compared to normal tissue. They suggested that
the lack of these vitamins caused the lighter coloring observed in
aposymbiotic specimens (66). Moreover, the authors claimed that

symbionts isolated from g. americana synthesized the three vitamins _ig

13

21559 (66). Unfortunately, the isolation techniques used by Ludwig and
Gallagher are subject to the criticisms of Brooks and Richards (17;
see above), and therefore the results should not be accepted as fact
without further investigation. The accumulated evidence strongly
suggests, however, that mycetocyte symbionts can be crucial to
cockroach development.

In contrast to the controversial studies regarding cockroach
symbionts, studies with a number of beetle species have proven quite
successful in defining roles of microbial symbionts in insect host
nutrition. Blewett and Fraenkel (9) and Pant and Fraenkel (81), in

extensive studies of larvae of the beetles Stegobium. paniceum and

 

Lasioderma serricorne, showed that intracellular symbionts supplied up

 

to six pritamins required by their hosts. Pant and Fraenkel’s (81)
study represents one of a very few wherein the symbionts, in this case
yeasts, were isolated and cultured, and subsequently introduced into
sterile (i.e., symbiont-free) host insects. The effects of the
mdcrobes and their metabolites on the hosts could therefore be clearly
established. In the absence of the yeast symbionts, the beetles
exhibited requirements for vitamins of the B-complex (9,81). When
aposymbiotic larvae inoculated with cultures of the yeast acquired
their normal symbiont flora, the vitamins were no longer required for
the insects’ growth and development (81). FUrthermore, a requirement
for cholesterol by the symbiont-free beetles was overcome by
inoculation of the insects with symbionts (81). It appears that the
microbes provide vitamins and cholesterol, or their precursors, to the
beetle host.

Pant and Fraenkel (82) also studied symbiosis in the beetle

l4

Oryzaephilus surinamensis L. Larval specimens incubated at elevated

 

temperatures appeared to lose their mycetocyte symbionts (82) . Growth
of the symbiont—free larvae was retarded, and, as with the species
previoisly investigated (see above), the symbionts were implicated in
the synthesis of B-vitamins for their host (82). Although the
symbionts were not isolated, and the studies were less extensive than
these authors’ previois work, the roles of the symbionts appear to be
similar. Pant and Fraenkel (81) suggested that vitamins produced by
the symbionts within mycetocytes become available to the host when the
insects digest their yeast symbionts, or following diffusion of the
vitamins into the hemolymph. The latter route was proposed as the more
likely (82) in view of Graebner’s (40) report of vitamin production and
secretion by pure cultures of the Stgobium symbionts.

The beetles studied by Pant and Fraenkel thrive on relatively
nutrient deficient diets (e.g., flour, leaves, herbs), supported by
their associations with vitamin-producing yeasts. Similarly,
blood-sucking insects require supplementation of their B-vitamin
deficient diet with products elaborated by gut bacterial symbionts
(4,10,18,89,105). In experiments similar to those described above,
Brecher and Wigglesworth (10) and Baines (4) demonstrated that a

bacterium, Nocardia rhodnii, isolated from guts of Rhodnius prolixus,

 

 

could supply its host with vitamins of the B—complex that were
necessary for the insect’s growth. Symbiont—free insects showed
retarded and incomplete development, whereas supplementation of their
diet with mixtures of B—vitamins could overcome the growth lag (4,10).
Furthermore, infection of sterile specimens with N. rhodnii restored

the insects’ ability to complete development at the same rate, and on

15

the same vitamin-poor diet, as normal insects (4) . Clearly, these
symbionts can have a significant influence in supplementing an
inadequate diet with essential vitamins.

For insects whose diets consist of N-poor wood or cellulose,
microbial symbionts may also have a critical role in N nutrition.
Crthocercus utilizes the microbes harbored in its hindgut as a soirce
of protein (27) . The microorganisms are apparently regurgitated into
the midgut from the hindgut and digested by their host (27). A similar
fate was observed for the cellulose-digesting bacteria associated with
Lamellicorn beetles (104), as described above. Tbth (96) suggested
that termites might also obtain microbial protein by digesting their
own microbiota. He surmised that microbes from the overcrowded hindgut
would be forced into the midgut (i.e., regurgitated), and there be
digested. However, studies by Kovoor (60) showed that the enteric
valve prevents food reflux from the hindgut to the midgut in termites.
Gut symbionts could not, therefore, be regurgitated for a source of
protein for the termite.

Termites have apparently evolved an alternate means by which to
utilize some of their gut microbial protein. The lower wood-eating
termites frequently partake in proctodeal feeding, i.e. , ingestion of a
microbe-laden droplet of the paunch (hindgut) contents of one termite
(donor) by another member of the colony (solicitor) (80). By this
practice, young larvae and newly molted termites are infected or
reinfected, respectively, with their necessary symbionts. Althoigh
some of the microbes survive the passage throigh the midgut to
inoculate the paunch, many of the protozoa lyse, and the protein
thereby released can be used by the termite. The significance of this

16

microbial protein to the overall N economy of termites does not appear
to be great, l'owever (46); additional N sources must also be
available.

Nitrogen Economy g Termites

 

The ability of xylophagous termites to thrive on N-poor diets of
sound wood or cellulose suggests that these insects have evolved an
efficient means to acquire and/or conserve N. The diet of termites
consists of plant material in various forms. Most of the lower
termites feed on dead wood, which may be sound or in various stages of
fungal decay (65,106). Sound, woody plant tissues contain only about
0.03 to 0.10% N by weight (34,49,63) and a very high carbon to nitrogen
ratio of up to 500:1 (34). The wood N is primarily in the form of
proteins, free amino acids and peptides, nucleic acids, and alkaloids
(63). Since termites excrete very small amounts of N in their feces
(49,65,106), their ability to utilize most of the wood nitrogenous
compounds must be acknowledged. Even so, the daily N intake would be
very low for most termites. For example, wood consumption by termites
in the family Rhinotermitidae is about 20 mg/(g fresh termite x day)
(42,106). If the wood consumed contained as much as 0.10% N by weight,
the total N intake would be only 0.02 mg/(g fresh termite x day).
Although the N needs of termites in their natural environment are not
documented, this amount of N would probably not suffice in itself for
colony growth and development, particularly during periods of swarming
(79).

Mauldin and Smythe (70) discussed the probable sources fbr N
acquisition by termites. Termite N could derive from: (1) dietary N,

from ingested food or digestion of symbionts; (2) microbial-mediated

17

fixation of atmospheric N and (3) utilization of microbial or, as

2.
LaFage and Nutting (63) suggest, termite nitrogenous waste products
(70). As mentioned above, digestion of symbionts may not offer
termites a significant source of N, although the extent of proctodeal
feeding by these social insects has not been documented. The symbiotic
protozoa of lower termites apparently divide only at post-molt periods
(3,56) when they must repopulate the termite hindgut. Moreover, Trager
(100) and Hungate (50,53) concluded that individual protozoa live for
relatively long periods. Protozoa would not, therefore, provide a
ready source of N for the termite by death and lysis.

Nitrogen from dietary sources other than sound wood could be
provided by fungi associated with the wood, or from the cannibalistic
and coprophagous habits of termites (63). Many termites prefer wood
that is at least partially degraded by certain fungi (65,90) , but
preferences vary amoung termite species and with different woods and
fungi (6,81). Hungate (49) investigated the potential contributions of

certain fungi to the N nutrition of lower termites. Working primarily

with Zootermqasis, Hungate found that some fungus-degraded woods

 

supported termite growth and N acquisition whereas others did not (49) .
When he analyzed the roles of fungi in termite N nutrition, he found
that fungal activities actually increased the N content of the wood
subsequently consumed by the termites (49) . Not only did fungi degrade
non-nitrogenous components of the wood, thereby decreasing the carbon
to nitrogen ratio, but they also transferred N from the surrounding
soil into the mod (49) .

Hendee (44) reported a beneficial effect of fungi on termite

growth and survival. Zootermopsis angusticolis exhibited better

 

18

survival, growth, development, and N acquisition on diets of
fungus-degraded woods than on sound woods. Hendee suggested that the
fungi offered a source of proteins and vitamins for the termites, and
extracellular enzymes from the fungi made wood more suitable for the
termite to digest. These conclusions correlated with the results of
Hungate (see above) and of Cleveland (24) , and have been supported by
more recent studies (6,59,90) . Moreover, consumption of

fungus-degraded wood by Reticulitermes flavipes altered the termites’

 

amino acid (21) and fatty acid (20) composition, suggesting differences
in the nutritive value of sound wood and degraded wood (21) . Finally,
fungi may play a role in detoxifying certain wood components harmful to
termites (6,44,90). The ultimate contribution of fungi to termite N
nutrition appears to vary depending upon the specific interactions
involved for a given termite species (e.g., see 90), but fungi could
certainly be important to the insect.

The N content of some termites’ diets may directly affect their
cannibalistic tendencies. Hendee (44) observed increased cannibalism

by g. angusticolis termites fed low N diets of sound wood compared to

 

termites fed fungus-degraded wood. Similarly, termites fed essentially
N-free diets exhibited high rates of cannibalism (2,33) . Even termites
fed presumably adequate (i.e., natural) diets practiced a regular
degree of cannibalism (44), suggesting that cannibalism occurs in the
termites’ natural habitat. Furthermore, chitinase activity has been
found in extracts of termites (98,102) . This enzyme may serve a role
in recycling N ingested with exuviae and cannibalized termites (72) .
The consequence of termite cannibalistic behavior may concern not only

N (and other nutrients) conservation, but also nest sanitation (6S) and

19

population control (63). Since no quantitative data regarding the
extent of cannibalism in natural colonies are available, it is
difficult to assess the contribution of this behavior to overall
termite N economy.

Coprophagy (i.e., ingestion of fecal material) may contribute to N
conservation in termite colonies by mediating a recycling of excreted
N. However, the significance of coprophagy to the termite is
questionable in view of the low N content of feces. Nitrogen
constitutes less than 0.25% of the feces of Zootermopsis (49), and

 

likewise a small percentage of Nasutitermes exitiosus excreta (65);

 

feces of other termites are probably colparable (106) . Uric acid, a
presumed waste product of termites (72) , constitutes an even lesser
portion of the excreta. Immgate (49) reported that fecal pellets of
Zcotermopsis contained only 0.02% uric acid, which accointed for less
that 10% of the total N excreted. Since the N content of the termites’
diet is so low, these data are not surprising. The fact may, however,
reflect the termites’ potential to efficiently recycle and conserve
their N internally, as suggested by leach and Granovsky (64) and as
discussed further below.

The roles of termite gut symbionts in dietary-related N
acquisition have not been clearly established (see 63). Speck et a1

(95) found that removal of gut bacteria from Nasutitermes nigriceps did

 

not affect the termites’ amino acid metabolism. mwever,
Reticulitermes santonensis specimens freed of their protozoa, bacteria,
or both, slowed reduced incorporation of l4C-glucose into amino acids

(95) . By contrast, removal of protozoa and most bacteria from guts of

ggptotermes formosanus did not alter the insects’ protein-bound amino

20

acid content (69,70). Further studies with Q. formosanus revealed

 

that their content of free amino acids was reduced by the removal of
gut bacteria and two of three major species of gut protozoa (69) .

Similarly, incorporation of 14

C-acetate into protein-bound amino
acids was reduced in the absence of the symbionts (69) .

Cleveland (26) postulated that fixation of atmospheric N2 coild
be important to termites. Althoagh termites survived when fed N-free
diets of filter paper, Cleveland was unable to detect N uptake by
termites when he used manometric techniques (26). Likewise, Hungate

(49,51) fomd no evidence for N fixatim by termites maintained am a

2
variety of wood, filter paper, or cotton diets. Greene and Breazeale
(41) reported the isolation of Nz-fixing bacteria from termites, but
did not detail their studies nor pursue their findings.

Tbth (96,97) investigated a number of insects, including termites
and aphids, for N2 fixation by using a “surviving system" technique.
He macerated and incubated insect tissue in a N-free organic broth,
then measured the increase in comined N in the preparation. Mole
termites assayed in this manner caused a 60% increase in the N content
of the medium (96,97) . Toth susequently isolated bacteria which
similarly appeared to bind N (97). However, the significance of
T6th’s results are questionable. Symbiont-free abdominal and thoracic
tissues of Kalotermes flavicolis appeared to incorporate more N than
did the symbiont-laden hindgut preparations (96) . Moreover, the
'symbimts' were isolated on medium which contained agar (97) , and
therefore was not completely free of colbined N. The organisms
isolated were not identified or quantitated, nor was their origin from

the termite hindgut firmly established (97). The importance of the

21

cultured bacteria to termite N2 fixation could not be determined.
Toth’s surviving system method for quantitating N2 fixation was
criticized by Smith (94). Using aphids, which Toth had also reported

as being capable of N fixation, Smith (94) measured the ability of

2
insects to take up 15N. No incorporation of 15N by the specimens
was detected (94). The sensitivity of Smith’s technique would have
allowed detection of N uptake equal to 0.03% of the insect’s total N
(94), a value much lower than T6th’s reported aphid N-binding activity
(97). Toth’s results with aphids were, therefore, refuted, and his
data for other insects assayed by the surviving system technique were
questionable. Adthough Smith did not assay termites for 15N
incorporation, the doubt cast on Toth’s work, in addition to Toth’s
own questionable results with termite tissues and "symbionts",
suggested that further studies of termite N2 fixation were necessary.
The potential role of symbiotic gut bacteria in termite N2
fixation has recently been elucidated (7,12,39,86). The sensitivity of
the acetylene reduction assay as a measure of N2 fixation (85) has
permitted quantitation of nitrogenase activity in numerous termite
species (7,11,12) and has revealed the potential significance of N2
fixation to termite development (11). The level of N2 fixation

exhibited by young larvae of Coptotermes formosanus could enable these

 

termites to double their protoplasmic N content in the period of one
year (11). Moreover, Breznak et a1 (12) clearly established that gut
bacteria mediated N2 fixation in termites. These workers found that
antibacterial drugs fed to termites abolished the insects’ NZ-fixing
ability (12). The loss was correlated with the loss of hindgut

bacteria, whereas protozoa were not directly affected (12). In

22

addition, the level of N2 fixation observed reflected the N content
of the termites’ diet: on a low N diet of filter paper, Nz—fixing
activities were relatively high, compared to lower activities on high N
diets (12).

Bacteria mediating N2 fixation in termites have been isolated
and identified (39,86). Potrikus and Breznak (86; see Appendix II)

isolated and quantitated NZ-fixing Emterobacter agglomerans from guts

 

of Q. formosanus. The bacteria, which fixed N2 only under anaerobic

 

conditions [the presumed state of the termite hindgut (8,101)], were

present in populations of about 200 cells/gut, or about 2.9 x 105

 

cell/ml of gut fluid (86) . The Nz-fixing activity of E. agglomerans

could be important to the N economy of _C_. formosanus (86).

 

NZ-fixing Citrobacter freundii were isolated, though not

 

quantitatively, from three species of Australian termites (39) , whereas

Enterobacter sp. capable of growth on N-free medium were isolated from

 

a number of other termite species (36) . The former isolates fixed N2
only under anaerobic conditions (39) whereas the latter appeared to be
active aerobically (36) . However, the medium or which the latter

Eaterobacter isolates were cultivated contained agar (36), and

 

therefore contained combined N. The authors noted the small size of
colonies on the isolation medium (36), suggesting unfavorable growth
conditions for the bacteria. It is reasonable to assume that the

Enterobacter isolates obtained by Eitick et al (36) were growing

 

aerobically by scavenging the small amounts of N contained in the
medium, rather than by fixing atmospheric N2.
In spite of confirmation of the _12 vivo ability of termite

symbionts to fix atmospheric N2, and the potential significance of

23

this activity, not all termites demonstrate high levels of N2
fixation (7,12). It would seem, therefore, that termites might have
yet another means to obtain or conserve N. Indeed, conservation of
colbined N could be very important to theN eoonomyof a termite
colony.

Leach and Granovsky (64) postulated that termite gut symbionts
conserve N by catabolizing urates or other nitrogenon wastes
elaborated into the hindgut by the insect’s Malpighian tubules. If the
N was incorporated into microbial protoplasm, it could be recycled to
the termite by proctodeal feeding or by direct absorption of
nitrogenous coupounds released from lysed microbes within the paunch
(64) . Although studies already discussed have shown that few protozoa
die within the termite hindgut, the potential for microbial-mediated N
conservation is significant. Protozoan and bacterial metabolites
released into the hindgut could be assimilated by the termite directly,
as has been hypothesized for the B-cotplex vitamins synthesized by
yeast symbionts of sore beetles (see above). '

ngate (49) considered the possibility of microbial recycling of
urates, but foind that 0.02% of the fecal pellet dry weight of

Zootermopsis was uric acid. Thus, he concluded that protozoa could not

 

efficiently metabolize uric acid, and N was therefore accreted and not
conserved (49). However, in view of the large mounts of uric acid
excreted by truly uricotelic insects (e.g., see 88), the small amount
of the purine in termite excreta appears insignificant.

In contrast to ngate’s opinion, Toth (97) felt that termite gut
symbimts corld indeed conserve N by catabolism of uric acid. He

reported both uricolytic and ureolytic activity in cultures of termite

24

bacteria incubated in the presence of uric acid and urea, respectively
(97) . Unfortunately, these microbes were neither identified nor
quantitated, and, as with Toth’s purported Nz-fixing termite
symbionts, their roles in the insect could not be affirmed. Sebald
(93) reported the isolatiom of uricolytic bacteria from guts of 5.
luciffls, but did rot study their roles in termite nutrition. Noirot
and Noirot-Timothée (80) suggested that bacteria in the mixed segment
of higher termites might reutilize nitrogenous wastes. In
Cephalotermes rectangularis, urine apparently passes from the
Malpighian tubules directly into the mixed segment, where an abundant
population of bacteria resides (80). No studies of these bacteria have
yet been reported.

Jucci (S4) and Moore (72) detected uric acid in termite specimens.
However, no quantitative measurements of uric acid in termites were
reported, nor was the potential significance of uric acid to termite N
nutrition investigated. From the meager evidence presented by Hungate
(49) regarding uric acid excretion by termites and from Jucci (54) and
Moore (72) regarding detection of uric acid in termites, Moore (72)
concluded that uric acid is a nitrogenous excretory waste of termites.
Moore (72) as well as Hungate (49) contended that uric acid could not
be utilized by termites, either directly or by intervention of
microbial symbionts, and therefore the purine was not important to

termite N conservation

STATEMENT CF PURPCBE

Since N acquisition is a problem critical to termite growth and
survival, it seemed wise to reconsider the possible significance of
microbial-mediated recycling of termite nitrogenous wastes. The
paucity of data regarding synthesis, storage, and excretion of
nitrogenous compounds by termites severely limits our understanding of
termite N nutrition. The potential significance of microbes in
conserving termite "excretory" N could not be investigated without more
knowledge of termite excretory products and processes. Studies with
other insects, particularly cockroaches, to which termites are
phylogenetically related (27) , have shown that uric acid can serve as
an important nitrogenous reserve (see above). The reports of Jucci
(54) and more (72) , who demonstrated that termites do contain uric
acid, and the lack of significant mounts of uric acid in termite feces
(49) , buttressed the contention that uric acid might serve as a
nitrogenous reserve material in termites, as it does in cockroaches.
Thus, it was felt that a quantitative investigation of the uric acid
coutent of termites and their excreta was necessary as a prelude to
investigating the roles of termite symbionts in the N metabolism of
their host.

The hypothesized roles of cockroach mycetocyte bacteria in uric

25

26

acid catabolism have been difficult to verify (see above). Termites,
excepting the species Mastotermes darwiniensis, do not possess
mycetocyte symbionts (55). Rather, an abundant population of protozoa
and bacteria reside in the readily accessible hindgut (ll,36,46,9l,93) .
Sebald’s (93) isolation of termite hindgut bacteria capable of
fermenting uric acid suggested that the role of termite symbionts in
uric acid turnover might be more easily addressed than that of the
uncultured cockroach bacteroids. Moreover, preliminary studies with

the termite Reticulitermes flagpes revealed significant populations of

 

putative uricolytic bacteria in the insect’s hindgut. The microbes
grew readily o‘u comon laboratory media supplemented with uric acid
(see Article II). These results prorpted the present investigation of
microbial intervention in termite uric acid metabolism and N
conservation.

The studies reported here were undertaken to investigate whether
the gut microbes of lower xylophagous termites could recycle N within
the insect by catabolizing the termite’s nitrogenous wastes. The
investigatious embrace a detailed, quantitative study of uric acid in
termites and their excreta; quantitation and identification of
uricolytic bacteria from guts of _13. flaviEs; studies of the
metabolism of uric acid (and other purines), including quantitative
analyses of the end products of uric acid fermentation, by the isolated
symbionts E 1i_t_r_q; measurements of in_ fig uricolysis by termite gut
bacteria; and determinations of the termite’s ability to synthesize
and transport uric acid, as well as to use the products of gut
microbial uricolysis. An understanding of the importance of microbial

uricolysis to termite N nutrition may help clarify a problem which has

2'7

puzzled biologists for many years, and advance our understanding of the
biology of termites. This intriguing insect-microbe symbiosis reveals
the potential significance of symbionts to uric acid catabolism in

higher organisms .

10.

LITERATURE CITED

Abo—Khatwa, N. 1978. Cellulase of fungus-growing termites: a
new hypothesis on its origin. Eacperientia 215559-560.

Andrew, B. J. 1930. Methods and rate of protozoan refaunation
in the termite Temsis -_a_umsticolis Hagen. Univ. Calif.
(Berkeley) Publ. Zool. 3_3:449-470.

Andrew, B. J., and S. F. Light. 1929. Natural and artificial
production of the so-called "mitotic flares” in the intestinal
flagellates of Temsis angusticolis. Univ. Calif.
(Berkeley) Publ. 2001. gun-44m.

 

 

Baines, S. 1956. The role of the symbiotic bacteria in the
nutritiou of Rhodnius prolixus (Hemiptera). J. exp. Biol.
33:533-541.

deBary, A. 1879. Die Erscheinung der miose. Trubner,
Strassburg.

Becker , G. 1971. Physiological influences on wood-destroying
insects of wood compounds and substances produced by
microorganisms. Wood Sci. Tech. 5:236-246.

Benemann, J. R. 1973. Nitrogen fixation in termites. Science
181:164-165.

Bignell, D. E., and J. M. Anderson. 1980. Determination of pH
and oxygen status in the guts of lower and higher termites. J.
Insect Physiol. 23:183-188.

Blewitt, M., and G. Fraenkel. 1944. Intracellular symbiosis and
vitamin requirements of two insects, Lasioderma serricorne and
Sitodr pa_nicea. Proc. my. Soc. London, Ser. B,
133: 12-221.

 

Brecher, G., and V. B. Wigglesworth. 1944. The transmission of
Actinoryces rhodnii Erikson in Rhodnius prolixus Stal
(Hemiptera) and its influence on the growth of the host.
Parasitol. 15:220-224.

28

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

29

Breznak, J. A. 1975. Symbiotic relationships between termites
and their intestinal microbiota, p. 559-580. In_ D. H.
Jennings and D. L. lee (eds.), Symbiosis (Society for
Experimental Biology symposium Ser., no. 29). Cmbridge Univ.
Press, Cmbridge.

Breznak, J. A., W. J. Brill, J. W. Mertins, and H. C.
Coppel. 1973. Nitrogen fixatiou in termites. Nature
_2_4_4:577-580.

Brooks, M. A. 1963. Symbiosis and aposymbiosis in arthropods.
Symp. Soc. gen. Microbiol. 13:200-231.

Brooks, M. A., and W. B. Kringen. 1972. Po1ypeptides and
proteins as growth factors for aposymbiotic Blattella germanica
(L.). p. 353-364. _IEJ. G. Rodriguez (ed.), Insect and mite
nutrition. Amsterdam: North-Holland.

Brooks, M. A., and A. G. Richards. 1955. Intracellular
symbiosis, in cockroaches. I . Production of aposymbiotic
cockroaches. Biol. Bull. _1_0_9_:22-39.

Brooks, M. A. , and A. G. Richards. 1956. Intracellular
symbiosis in cockroaches. III. Rte-infection of aposymbiotic
cockroaches with symbiotes. J. exp. 2001. 13:447-465.

Brooks, M. A., and K. Richards. 1966. m the in vitro culture
of intracelluar symbiotes of cockroaches. J. Invert. Pathol.
8:150-157.

Buchner, P. 1965. Ehdosymbiosis of animals with plant
microorganisms. Wiley-Interscience, New York.

Bursell, E. 1967. The excretion of nitrogen in insects. Adv.
Insect Physiol. 4:33-67.

Carter, F. L., L. A. Dinns, and R. V. Smythe. 1972. Effect
of wood decayed by lenzites trabea ou the fatty acid
. composition of the eastern subterranean termite, Reticulitermes
flaviE . J. Insect Physiol. _Jé:1387-1393.

 

Carter, F. L., and R. V. Smythe. 1973. Effect of sound and
Linzentes—degraded wood ou the amino acid corposition of
Reticulitermes flaviEs. J. Insect Physiol. 12:1623-1629.

Chapuen, R. F. 1975. The insects: structure and function.
American Elsevier Publishing Co., Inc. New York.

Cleveland, L. R. 1923 . Symbiosis between termites and their
intestinal protozoa. Proc. Natl. Acad. Sci. USA
_9_1 : 424-428 .

Cleveland, L. R. 1924. The physiological and symbiotic
relationships between the intestinal protozoa of termites and
their host. Biol. Bull. 56:178-225. -

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

30

Cleveland, L. R. 1925. The effects of oxygenation and
starvation ou the symbiosis between the termite, m, and
its intestinal flagellates. Biol. Bull. 4_8:309-327.

Cleveland, L. R. 1925. The ability of termites to live perhaps
indefinitely on a diet of pure cellulose. Biol. Bull.
4_8_:289-293.

Cleveland, L. R., S. R. Hall, E. P. Sanders, andJ. Collier.
1934. The wood-feeding roach Cryptocercus, its protozoa, and
the symbiosis between protozoa and roach. Mem. Am. Acad.
Arts Sci. gum—342.

 

Cochran, D. G. 1973. Colparative analysis of excreta from
twenty cockroach species. Corp. Biochem. Physiol.
5_6_A:409-419.

Cochran, D. G. 1975. accretion in insects, p. 178-281. E D.
J. Candy and B. A. Kilby (eds.), Insect biochemistry and
function. Chapman and Hall, London.

Cochran, D. G. 1976. Excreta analysis on additioual cockroach
species and the louse cricket. Corp. Biochem. Physiol.
53A:79-81.

Cochran, D. G. 1979. Corparative analysis of excreta and fat
body from the various cockroach species. Corp. Biochem.

Cochran, D. G., D. E. Mullins, and K. J. Mullins. 1979.
Cytological changes in the fat body of the American cockroach,
Per iplaneta mer icana, in relation to dietary nitrogen levels.
Ann. Eutolol. Soc. Amer. 2_2_:197-205.

Cook, 8. F., and K. G. Scott. 1933. The nutritional
requirements of Zootermcpsis (Termrpsis) angusticolis. J.
Cell . Corp. Physiol . 5395-110 .

Cowling, E. B., and W. Merrill. 1966. Nitrogen in wood and its
role in wood deterioration. Can. J. Bot. 4_4_:1539-1554.

Donnellan, J. F., and B. A. Kilby. 1967. Uric acid metabolism
by symbiotic bacteria from the fat body of Periplaneta
americana. Corp. Biochem. Physiol. 2_2_:235-252.

Eutick, M. L., R. W. O’Brien, and M. Slaytor. 1978. Bacteria
from the gut of Australian termites. Appl. Environ.
Microbiol. 3_5:823-828.

Franke-Grosmann, H. 1967. Ectosymbiosis in wood-inhabiting
insects, p. 141-205. lugs. M. Henry (ed.), Symbiosis, vol.
2. Academic Press, NewYork.

French, J. R. J. 1975. The role of termite hindgut bacteria in
wood decorposition. Mater. Org. l_0:1-13.

39I

40.

41.

42.

43.

44.

45.

46.

47.

48.

43.

50.

51.

52.

53.

31

French, J. R. J., G. L. Turner, and J. F. Bradbury. 1976.
Nitrogen fixation by bacteria from the hindgut of termites. J.
Gen. Microbiol . 25: 202-206 .

Graebner, K. E. 1954. Vergleich Morphologische und
Physiologische Studien an Anobi iden und Cerambycidensymbioten.
z. Morph. Okol. Tiere, 4_1_:471-523.

Greene, R. A., and E. L. Breazeale. 1937. Bacteria and the
nitrogen metabolism of termites. J. Bacteriol. 33:95-96.

Haverty, M. I. 1976. Termites. Pest Control 14:12-17, 46, 47,
49.

Haydek, M. H. 1953. Influence of the protein level of the diet
on the longevity of cockroaches. Ann. Entomol. Soc. Amer.
46:547-560.

Hendee, E. C. 1935. The role of fungi in the diet of the common
dampwocd termite, Zootermopsis angusticolis. Hilgardia
9:499-525.

 

 

Henry, S. M., and R. J. Block. 1962. Amino acid synthesis, a
ruminant like effect of the intracellular symbionts of the
German cockroach. Fed. Proc. almbstr. 9.

Honigberg, B. M. 1970. Protozoa associated with termites and
their role in digestion, p. 1-36. Ink. Krishna and F. M.
Weesner (eds.), Biology of termites, vol. 2. Academic Press,
New York.

Hungate, R. E. 1938. Studies on the nutrition of Zootermopsis.
II. The relative importance of the termite and the protozoa in
wood digestion. Ecology, 1931-25.

 

Hungate, R. E. 1939. Ebrperiments on the nutrition of
Zootermopsis. III. The anaerobic carbohydrate dissimilation
by the intestinal protozoa. Ecology, 2_0:230-245.

 

Hungate, R. E. 1941. Experiments on the nitrogen economy of
termites. Ann. Entoml. Soc. Amer. 3_4:467-489.

Hungate, R. E. 1943. Quantitative analyses on the cellulose
fermentation by termite protozoa. Ann. Eutorol. Soc. Amer.
_Ifi:730-739.

Hungate, R. E. 1944. Termite growth and nitrogen utilization in
laboratory cultures. Texas Acad. Sci. Proc. Trans. _2_7_:1-7.

Hungate, R. E. 1946. The symbiotic utilization of cellulose.
J. Elisha Mitchell Sci. Soc. _6_3:9-24.

Hungate, R. E. 1955. Mutualistic intestinal protozoa, p.
159-199. In S. H. Hutner and A. Lwoff (eds.) , Biochemistry
and physiology of protozoa, vol. 2. Academic Press, New York.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

32

Jucci, C. 1921. Sulla presenza di depositi uratici nei tessuto
adipose dei termitidi. Atti. Accad. naz. Lincei Memorie
39:213-215.

Jucci, C. 1952. Symbiosis and phylogenesis in the Isoptera.
Nature, 1_6_9_:837.

Kirby, H. 1941. Protozoa and other animals, p. 890-1008. I__n G.
N. Calkins and F. M. Summers (eds. ), Protozoa in biological
research. Columbia Univ. Press, New York.

Koch, A. 1956. The experimental elimination of symbionts and its
consequences. Exptl. Parasitol. 5: 481-518.

Koch, A. 1967. Insects and their endosymbionts, p. 1-106. In
S. M. Henry (ed. ), Mbiosis, vol. 2. Academic Press, New
York.

Kovoor, J. 1964. Modifications chimiques provoquées par un
Termitidé (Microcerotermes edentatus, Was.) dan du boid de
peuplier sain ou partiellerent dégradé par des champignons.
Bull. Biol. Fr. Belg. 98:491-510.

 

Kovoor, J. 1967. Etude radiographique du transit intestinal chez
un termite supérieur. Ebcperientia, 23:820-821.

Kovoor, J. 1970. Presence d’enzymes cellulolytiques dans
l’intestin d’un termite supérieur, Microcerotermes edentatus
(Was.). Ann. Sci. Nat. Zool. Biol. Anim. 13:65-71.

 

Krishna, K. 1969. Introduction, p. 1-17. 3} K. Krishna and F.
M. Weesner (eds.) , Biology of termites, vol. 1. Academic
Press, New York.

LaFage, J. P., and w. L. Nutting. 1978. Nutrient dynamics of
termites, p. 165-232. In M. V. Brian (ed.), Production
ecology of ants and termites. Cambridge Univ. Press,
Cambridge.

leach, J. G., and A. A. Granovsky. 1938. Nitrogen in the
nutrition of termites. Science, 8_7:66-67.

Lee, K. E., and T. G. Wood. 1971. Termites and soils.
Academic Press, New York.

Ludwig, D., and M. R. Gallagher. 1966. Vitamin synthesis by
the symbionts in the fat body of the cockroach, Per iplaneta
mericana (L.). J. N. Y. EntoIol. Soc. 24:134-139.

 

Maddrell, S. H. P. 1971. The mechanisms of insect excretory
systems, p. 199-331. 111 J. W. L. Bement, J. E.
Treherne, and V. B. Wigglesworth (eds.) , Advances in insect
physiology, vol. 8. Academic Press, New York.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

33

Martin, M. M., and J. S. Martin. 1978. Cellulose digestion in
the midgut of the fungus-growing termite Macrotermes

natalensis: the role of acquired digestive enzymes. Science,
199:1463-1455.

 

Mauldin, J. K., N. M. R' h, and D. W. Cook. 1978. Amino
acid synthesis from C-acetate by normally and abnormally
faunated termites, Coptotermes formosanus. Insect Biochem.

Mauldin, J. K., and R. V. Smythe. 1973. Protein-bound amino
acid content of normally and abnormally faunated Formosan

termites, Coptotermes formosanus. J. Insect Physiol.
_Jfi:1955-1960.

 

 

Misra, J. N., and V. Ranganathan. 1954. Digestion of cellulose
by the mound building termite, Termes (cyclotermes) obesus
(Rambur) . Proc. Indian Acad. Sci. 32:100-113.

Moore, B. P. 1969. Biochemical studies in termites, p.
407-432. _1_r1 K. Krishna and F. M. Weesner (eds.) , Biology of
termites, vol. 1. Academic Press, New York.

Mullins, D. E. 1974. Nitrogen metabolism in the American
cockroach: an examination of whole body mmonium and other
cations excreted in relation to water requirements. J. exp.
Biol. 61:541-556.

Mullins, D. E., and D. G. Cochran. 1973. Nitrogenous
excretory materials from the American cockroach. J. Insect
Physiol. _1_9_:1007-1018.

Mullins, D. E., and D. G. Cochran. 1975. Nitrogen metabolism
in the American cockroach-I. An examination of positive
nitrogen balance with respect to uric acid stores. Comp.
Biochem. Physiol. _50_A:489-500.

Mullins, D. E., and D. G. Cochran. 1975. Nitrogen metabolism
in the American cockroach-II. An examination of negative
nitrogen balance with respect to mobilization of uric acid
stores. Comp. Biochem. Physiol. §0_A:501-510.

Mullins, D. E., and D. G. Cochran. 1976. A corparative study
of nitrogen excretion in twenty-three cockroach species. Comp.
Biochem. Physiol . S3_A: 393-399 .

Mullins, D. E., and C. B. Keil. 1980. Paternal investment of
urates in cockroaches. Nature, 2_81:567-569.

Nielson, M. G., and G. Josens. 1978. Production by ants and
termites, p. 45-53. _I_r_u_ M. V. Brian (ed.), Production
ecology of ants and termites. Cmbridge Univ. Press,
Cambridge.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

34

Noirot, C., and C. Noirot-Timothée. 1969. The digestive
system, p. 49-88. E K. Krishna and F. M. Weesner (eds.) ,
Biology of termites, vol. 1. Academic Press, New York.

Pant, N. C., and G. Fraenkel. 1954. Studies ou the symbiotic
yeasts of two insect species, Lasifirma serricorne F. and

Pant, N. C., and G. Fraenkel. 1954. On the function of the
intracellular symbionts of Oryzaephilus surinamensis L.
(Oucujidae, Coleoptera). J. Zool. Soc. India, 6:173-177.

Pierre, L. L. 1964. Uricase activity of isolated symbionts and
the aposymbiotic fat body of a cockroach. Nature, _20_1:54-55.

Pochon, J., H. deBarjac, and A. Roche. 1959. Recherches sur la
digestiou de la cellulose chez 1e termite Wrotermes
ghaerothorax. Ann. Inst. Pasteur, Paris 26:352-355.

 

Postgate, J. 1971. The acetylene test for nitrogenase, p.
311-315. E J. R. Postgate (ed.) , The chemistry and
biochemistry of nitrogen fixation. Plenum Press, New York.

Potrikus, C. J., and J. A. Breznak. 1977. Nitrogen-fixing
Enterobacter agglorerans isolated from guts of wood-eating
termites. Appl. Environ. Microbiol. 12 : 392-399.

Potts, R. C., and P. H. Hewitt. 1973. The distribution of
intestinal bacteria and cellulase activity in the harvester
termite Trinervitermes trinervoides (Nasutitermitinae) .
Insectes Sociaux, Paris 30:215-220.

Razet, P. 1966. Ies elements terminaux du catabolisme azoté
chez 1es insectes. Année Biol. 20:43-73.

Richards, A. G., and M. A. Brooks. 1958. Internal symbiosis
in insects. Ann. Rev. Eutorol. 1:37-56.

Sands, W. A. 1969. The association of termites and fungi, p.
495-524. 1}; K. Krishna and F. M. Weesner (eds.) , Biology of
termites, vol. 1. Academic Press, New York.

Schultz, J. E., and J. A. Breznak. 1978. Heterotrophic
bacteria present in hindguts of wood-eating termites
[Reticulitermes flavi s (Kollar)]. Appl. Environ.
Microbiol. _35:930-§36.

Schultz, J. E., and J. A. Breznak. 1979. Cross-feeding of
lactate between Streptococcus lactis and Bacteroides sp.
isolated from termite hindguts. Appl. Environ. Microbiol.
_3_7_: 1206-1210 .

 

Sebald, M. 1962. Etude sur les bactéries anaérobies
gram-negatives asporulées. Imprimerie Barnéoud S.A. , Laval.

94.

95.

96.

97.

98.

99.

100 .

101.

102.

103 .

104 .

105.

106.

107.

108 .

35

Smith, J. D. 1948. ambiotic micro-organisms of aphids and
fixation of atmospheric nitrogen. Nature, 16_2_:930-931.

Speck , U. , G. Becker , and M. Lenz . 1971.
Ernahrungsphysiolog ische mtersuchungen an Termiten nach
Selektiver Medikamentbser Ausschaltung der Darmsymbioten. z.
Angew. Zool. _S_8_:475-491.

Téth, L. 1948. Nitrogen-binding by Kalotermes flavicollis
(Isoptera) and its symbionts. Himg. Acta Biol. 1:22-29.

Tbth, L. 1952. The role of nitrogen-active microorganisms in
the nitrogen metabolism of insects. Tijdshrift Entoml.
95:43-62.

Tracey, M. V., and G. Youatt. 1958. Cellulose and chitinase in
two species of Australian termites. Emzymologia 12:70-72.

Trager, W. 1932. A cellulase from the symiotic intestinal
flagellates of termites and the roach Cryptocercus punctulatus.
Biochem. J. 291762-1771.

 

Trager, W. 1934. The cultivation of a cellulose-digesting
flagellate, Trichoionas termqusidis, and of certain other
termite protozoa. Biol. bull. 66:182-190.

 

Veivers, P. C., R. W. O’Brien, and M. Slaytor. 1980. The
redox state of the gut of termites. J. Insect Physiol.
26:75-77.

Waterhouse, D. F., R. H. Hackman, and J. W. McKellar. 1961.
An investigatiou of chitinase activity in cockroach and termite
extracts. J. Insect Physiol. 6:96-112.

Weesner, F'. M. 1960. Evolutiou and biology of termites. Ann.
Rev. Eutouol. _5_:153-l70.

Wiedemann, J. F. 1930. Die Zelluloseverdauung bei
Lalellicornierlarven. z. lbrphol. Okol. Tiere, 12:229-258.

Wigglesworth, V. B. 1972. The principles of insect physiology.
Chapman and Hall, Iondon.

mod, T. G. 1978. Food and feeding habits of termites, p.
55-80. 13 M. V. Brian (ed.) , Production ecology of ants and
termites. Cambridge Uhiv. Press, Cmbr idge.

Yamin, M. A. 1978. Axenic cultivation of the cellulolytic
flagellate Trichomitopgig termidis (Cleveland) from the
termite Zootermopsis. J. Protozool. §:535-538.

 

 

 

Yamin, M. A. 1980. Cellulose metabolism by the termite

flagellate Trichomigfiis termopsidis. Appl. Euviron.
Microbiol. 1_: 5 .

)I E‘ e .4
«- "L
r " - 2

r

I...
I
-'rn
A ‘r
tltfi' 21"
‘,V¢‘u.'
,1
A. ww‘

0'4:
..
1’
I
fi'l-

'cw
l

ARTICLE I

URIC ACID IN FUD-BATES W

C. J. Potrikus and J. A. Breznak

Reprinted from Insect Biochem. 10:19-27 (1980)

36

Insect Biochem.. Vol. 10. pp. 19 to 27.
@iPergamon Press Ltd. l980. Printed in Great Britain.

37

0020—1700 80 OHM-0019 $02.00 0

URIC ACID IN WOOD-EATING TERMITES“

C. J. PorRtxus and J. A. BREZNAK

Department of Microbiology and Public Health. Michigan State University. East Lansing. Ml 48824.
U.S.A.

(Received 9 April l979)

Abstract—Uric acid (UA) was present in termites as assessed by paper and thin layer chromatography. by
U.V. spectroscopy. by reactivity with uricasc. and by gas chromatography coupled to mass spectrometry.
Specimens of six species (Reticulitermes flavipes. R. virginicus. Coptotermes formosanus. Marginitermeu
huhhardi. Paraneotermes simplicicorm‘s. and C r_t'ptoterm¢'s cavifrons) contained UA in amounts accounting
for between l and 45",, of thc termttcs' dry weight and for between 4 and 69",, ofthc termitcs' total nitrogen.
Almost all ofthc UA in Rflavipes was associated with fat body tissue. Facccs contained less than 0.2"0 ofits
weight as UA. During 15 months ofcapttvtty‘. UA in Rflawpes workers increased from 17".. to 45.4“o ofthc
insects' dry weight at a linear rate of 2.7"n per month. Elemental analyses of termites during this period
indicated that synthesis of UA occurred. in part. at the expense of some endogenous nitrogen and carbon.
Compounds tentatively identified as inosinc and kynurentc acid were also present in termite extracts. Results
indicate that termites can store large quantities of UA internally. but they do not void the purine in
significant amounts despite the lack of detectable uricasc activity in termite tissues.
Key Word Index: uric acid. termite. nitrogen. fat body. uricasc. inosine. kynurcnic acid. fences

INTRODUCTION

THE ABILITY of many termites to thrive on nitrogen-
poor diets of sound wood or cellulose is intriguing in
terms of the insects‘ nitrogen-economy. HUNGATE
(1944) found that bound nitrogen in wood and soil,
when acted upon by fungi. was sufficient to support
good growth of laboratory colonies of Zootermopsis
species. However. recent studies (Portukus and
BREZNAK. 1977 and references therein) also support
the long-held belief that N 2 fixation by gut microbes
supplements the intake of dietary nitrogen by some
termites. LEACH and GRANOVSKY (I938) had advanced
an alternative hypothesis for nitrogen conservation by
termites. They postulated that the nitrogen in uric acid
(UA), a presumed waste product of termites, could be
recycled by the action of uricolytic gut bacteria.

Two crucial elements of the LEACH and GRANOVSKY
hypothesis are: (i) termites form UA (or a similar
compound); and (ii) termites contain uricolytic
organisms in their hindguts. Although there are a few
reports of UA formation by termites (Juccu. l92l;
HUNGATE, I941; MOORE. 1969). detailed and specific
chemical analyses are lacking. In this paper we provide
conclusive evidence for the presence of UA in termites
and estimate its concentration and distribution in
termite tissues and in faeccs. A separate communi-
cation (in preparation) will document the presence
of uricolytic bacteria in termite hindguts.

MATERIALS AND METHODS

Termite: and termite forces. Reticulitermes flavi'pes was
collected in Janesville. WI and Spring Arbor. MI and
C optotermesformosanus was collected in Lake Charles. LA.

 

‘ Journal article No. 887l from the Michigan Agricultural
Experiment Station.

flat‘lpt't

Both spears were maintained in the laboratory as previously
described (S('HL'LTZ and Bonus. I978). Specimens of R.
and R. l‘lrgllllt‘lfl from Mississippi. and
Paranuotcrmcs stmp/uicorms and Marginitermcu hubhardi
from Arizona. were gifts from Dr. M. l. HAVERTY. Samples of
(‘rrptnicrnm cart/mm collected in Florida were kindly
supplied by Dr. Svnxtu' TAM“.

Faccal pellets were obtained from galleries in infested
wood. To insure that extraneous (i.e. non-faccal) material
was not included for the analyses. only those pellets which
were distinctly prolatc spheroid in shape were selected. For
collection of fresh excreta. one hundred termites were placed
on a disc of stainless steel mesh screen fitted into a 60 ml jar.
After 12-24 hr the screen was removed and the adhering
facccs. which in this case consisted both of formed pellets and
pasty amorphous excrement. were scraped off and processed
immediately.

Unless indicated otherwise. all samples for analyses were

dried at 85C for 8 hr. pulverized with a porcelain pestle. and
then stored in a vacuum dessicator over CaSO‘ until required
for assay.
Extraction of uric acid. Routinely I-5 mg of powdered
termite tissue were suspended in 2 ml of 0.6". Li1CO3 for 10
min at 60 C. followed by ccntrifugation at 30003 for IS min
at 25 C. The supernatant fluid was then used immediately for
paper (PC) or thin layer (TLC) chromatography; for
determination of U.V. absorption spectra; or for enzymatic
assay of UA. Experiments showed that all the UA was
solubiltzcd by the one extraction.

Facccs were extracted as described above for TLC. but
were extracted with hot (85 C) distilled water for IS min for
enzymatic assay of UA. The latter procedure extracted all the
UA. but did not solubiltzc Li2C03-soluble. U.V.-absorbing
material in facces which interfered with the UA assay.
Paper and thin layer chromatographv. Samples were co-
chromatographcd with standards on sheets of Whatman No.
l qualitative filter paper or on DJ or 0.25 mm cellulose MN
300 TL plates (Brinkman) by using ascending solvents in one
or two dimensions. From over thirty solvent systems
investigated. the three yielding the mom satisfactory
separations in one dimension. and chosen for routine use.
were: (A) isopropanol: H20. 10:3 (vxv) (DIKSTEIN et al..

20 C. J. Poriuiws AND J. A. BREZNAK

I956); (B) n-butanolzacetic acidzH 10. 12:3:5 (by vol.)
(DONNI LLAN and Run. 1967); and (C) n-
propanol:NH.OH:H10. 60:26:” (by vol.) (PATAKI. I967).

Compounds were detected on chromatograms by viewing
under a short wavelength U .V. lamp (Mineralite. Ultra-Violet
Products. Inc.. San Gabriel. CA) andior by using the
diphenylcarbazone (DPC) spray reagent (DIKSTEIN et al..
I956). When desired. spots visualized by UV light were
scraped from TLC plates and eluted from the cellulose with
0.6“o Li,co,.

All standards for PC and TLC were prepared in 06",,
Li.CO .

Enemilitic assay of uric acid. UA content of termites and
faeces was estimated by measuring the decrease in
absorbance at 292 nm after treatment of extracts in 0.1 M
glycine buffer (pH 9.4) with 0.0l5 units of purified uricase
(hog liver. Type V; Sigma Chem. C0; Technical Bulletin No.
292-UV). Spectrophotometric measurements were made by
using a Varian model 6345 double-beam recording
spectrophotometer coupled to a Sargent-Welch model SR
recorder. All values for UA are reported as 2. dry weight.
unless indicated otherwise.

Gas chromatography and mass spectroscopy (CC-MS). For
GC—MS of UA. termites were extracted as described above
except that 5 M KOH was used. Tetraethyl derivatives of UA
were then prepared (ISMAIL and Ducts. I975). extracted into
diethyl ether. and separated by GC with a 1.83 m x 2 mm id.
glass column packed with 3"o (w ’w) OV-l7 on 80100 mesh
Supelcoport (Supelco. lnc.. Bellefonte. PA). Nitrogen was
the carrier gas (30 ml, min) and operating temperatures were:
column. 200 C; injector. 260 C; and flame ionization
detector. 270 C. Standards were commercial UA. derivatized
as described above. and caffeine. which served as an internal
reference.

Appropriate GC fractions were diverted for MS analysis to
an LKB 9000 mass spectrometer operating at an ionization
potential of 70 eV and a temperature of 250 C.

Total nitrogen and total carbon determinations. Total nitrogen
content of termites and faeces was determined by the micro-
Kjeldahl method of JOHNSON (l94l). except that a 6 hr
digestion time was used. Bovine serum albumin w d . used as a
standard. Total nitrogen and carbon were in addition
determined by using a C. Erba model I l04 elemental analyzer
equipped with a CSI model 208 digital integrator. The
combustion temperature was 1050 :t 20C and the internal

 

0350 111v

 

 

 

 

0225 250 275 300 325 350
h (nm)
Fig. l. U.V. absorption spectra of LizCO, extracts of
standard uric acid ( ) and of R. flavipes workers
( ------ ) before and after exposure to commercial hog liy er
uricase enzyme for 30 min.

 

standard was cyclohexanone-2.4-dinitrophenylhydrazone.
Finally. a few samples were commercially analyzed for total
nitrogen by Galbraith Laboratories. Inc.. Knoxville. TN.
Uricase assay. Whole termites. termite bodies with guts
removed. or termite guts were homogenized in 0.5 M
potassium phosphate buffer (pH 7.4) and assayed for uricase
activity by the method of SCHNEIDER and HOGENBOOM ( l 952).
Out homogenates were assayed immediately. whereas others
were first dialyzed at 4:C against 0.0l M Na borate buffer
(pH 10.0) for 36 hr to remove interfering levels of endogenous
UA. Mouse liver homogenate was used as a positive control
(SCHNEIDER and HOGENBOOM. 1952). Protein was determined
by the biuret method (HERBERT et al.. l97l) with bovine
serum albumin as standard.

 

lOO '-

50'-

RELATNE INTENSITY

 

 

280

252

224

237

 

 

)— 265

 

 

300

m/e
Fig. 2. Mass spectrum of tetraethyl uric acid prepared from workers of R. flat-tries.

38

Uric acid in termites 2|

N 1 fixation assay. Live termites were assayed for N 2 fixation
activity by methods previously described (BREZNAK et al..
1973).

Chemicals. UA. LizCOr bovine serum albumin and
chromatography standards were obtained from Sigma
Chemical Co.. St. Louis. MO. All other chemicals were of
reagent grade and were obtained from various commercial
sources.

RESULTS

Identification of uric acid in R. flavipes

U.V. absorption spectra of Li2C03 extracts of R.
flavipes workers were similar to that of standard UA.
exhibiting maxima at 292 and 236 nm and minima at
262 nm (Fig. 1). Treatment of such preparations with
uricase obliterated the 292 nm absorption maximum
(Fig. 1).

PC and TLC of R. flavipes extracts revealed the
presence of a prominent DPC-reactive, U.V. light-
absorbing compound whose migration rate was
similar to that of standard UA. RUA values for this
compound in solvent systems A, B, and C were 1.00,
and values ranged from 0.94 to 1.00 in all solvent
systems employed. Eluates of the termite compound
and standard UA possessed identical U.V. absorption
spectra. similar to those depicted in Fig. 1. Likewise,
the 292 nm absorption maximum of such eluates was
uricase-sensitive.

GC revealed two major derivatives (1 and 11) present
in termite extracts as well as in solutions of standard
UA. Retention times of these compounds relative to
the caffeine standard were: compound 1, 1.30-1.33;
compound 11. 2.15-2.22. Mass spectra of compounds 1
and 11 were virtually identical to each other and to that
reported for tetraethyl UA isomers (Fig. 2) (ISMAIL
and DAKIN. 1975). The molecular ion was observed at
mass/charge (m/e) = 280. Successive losses of 28 mass
units each, corresponding to loss of the ethyl groups as
ethylene. resulted in peaks at m/e = 252, 224, 196, and
168. Peaks at m/e = 265, 237, 209, and 181 indicated
loss of methyl groups which presumably originated
from the four ethyl ligands (lsuxtt and DAKIN, 1975).
From these data we concluded that termites contain
UA.

Uric acid content of R. flavipes worker termites

Freshly-collected R. flavipes workers contained
14.53;, UA. However, during laboratory captivity the
UA content increased markedly. Data for R. flavipes
population I (collected in Janesville, WI in June, 1977)
are given in Table 1. A more than twenty five-fold
increase in UA was observed over the 15 month period
of captivity. The relationship between UA content and
length of captivity was evaluated by regression
analysis and graphed by the method of least squares
(Fig. 3; population I). As can be seen, the rate of
increase of UA (i.e. the regression coefficient) was
approx. 2.7"/o per month. Data for workers from a
separate collection of R. flavipes (population 11;
Janesville, June, 1978) are also graphed in Fig. 3. The
regression coefficient for population 11 revealed that
UA increased at a rate of 2.1"/o per month during the

541 months of captivity. Samples of workers from a '

third collection of R. flavipes (Janesville; June, 1976)
as well as from a population of C. formosanus, initially

39

Table l. Uric acid (UA) and total nitrogen (TN) content of
R. flavipes workers during captivity“

 

Length of Average dry

 

. . . , °°UA-N
captivity weight per ‘3“ of dry weight __ x 100
(montth’termite (mgl‘ UA TN' °..TN
0 0.81 1.7 11.06 5.1
0.13 0.70 2.9 10.55 9.1
4 0.63 15.1 14.57 34.5
5 — 24.6 14.30 57.3
6 0.63 26.0 14.69 58.8
7.5 0.67 26.1 14.81 58 7
9 0.81 32 4 15.56 69.2
10.5 1.06 31.1 15.41 67.3
14 1.10 39.3 19.74 66.5
15 1.08 45.4 22.87 66.0

 

‘Population 1. Collected in Janesville. WI in June. 1977.

'Zero month refers to termites assayed within 24 hr of
collection. ,

‘Determined by drying a mass of termites to constant
weight and dividing this value by the number of termites in
the sample.

‘Mean of at least two determinations. Standard errors
were within 40".. (UA) or l.5"., (TN) of the mean.

'Values for commencement or between 4—6 months
captivity were determined with the C. Erba elemental
analyzer. Total nitrogen at 0.13 month was determined by
Kjeldahl analysis. Other values were determined by
Galbraith Laboratories. Inc.

’ Not determined.

contained 2‘}; UA which increased to 243;, after 7 and
14 months, respectively.

The total nitrogen content of population 1 workers
doubled during the 15 month period (Table 1).
Moreover, statistical analysis revealed that the UA
and total nitrogen content were highly correlated
(Pearson's r = 0.92). However, regression analysis of
total nitrogen values yielded the line shown in Fig. 3
(individual data points not depicted). The rate of
increase of total nitrogen was only 0.663;, per month. If
the increase in uric acid nitrogen (UA-N) occurred
soley at the expense of exogenous nitrogen. one might
expect the total nitrogen to increase at a rate one third
that of UA (i.e. 1/3 x 2.7 = 0.93; per month) since one
third UA is nitrogen by weight. However, the observed
rate of change of total nitrogen (i.e. 0.66?0 per month)
was significantly different from the expected rate at the

 

8

units: (tottcuntte dry wt.)
0
o

 

 

 

1° 1

’0 It

0 .. l l h
(l l 0 12 10

 

Month. Captivity

Fig. 3. Uric acid (UA) and total nitrogen (TN) content of

R. flavipes during laboratory captivity. C oefficients of

determination (r1) for the lines were 0.94 (H). 0.83
(I—I ). and 0.88 (--—).

22 C. J. PorRtkL's AND J. A. BREZNAK

Table 2. Uric aCid tliA) and total nitrogen (TN) content of various termite species,

 

 

 

Approx.
length of “I_ijAA'
captmty Developmental "0 dry weight“ 0 —'— x 100
Family Species Origin (months) Stage UA TN' ,oTlN
Rhinotermitidae Rflai-ipcs MissiSSippi Larva 3.5 9.35 12.3
Worker 1.8 1023 5.7
Nymph 3.0 7.97 12 7
Soldier 2.2 10.42 6.9
R. virgimcus Mississippi Larva 3.0 8.62 11.6
Worker 1.1 9 99 3.7
Nymph 1.2 7.23 5.4
Soldier 1.6 10.82 5.0
C. formosanus Louisiana Worker 4.7 10.22 15.2
Soldier 2.3 7.42 10.5
Alate 5.3 9.53 18.6
Kalotermitidae P. simplicimmie Arizona 60 Larva 7.8 8.34 31.2
Pseudergate 8.4 7.35 38.0
Nymph 5.1 _ 4 -
Soldier 4. 9 10,75 15.2
M. huhhardi‘ Arizona Pseudergate 13.6 -— —
Nymph 9.5 9.04 34.9
° Soldier 11.5 — —
C. cavi/rons Florida Worker 7.6 — —

 

‘Mean of at least two determinations. Standard errors were within 4. 0"

(UA) or 2. 0“ (TN) of the mean.

’Values were determined with the C. Erba elemental analyz.er

‘ UA content of faecal pellets of P. simplicicornis and M. hubbardi was < 0. 01“,, and <0. 04’.

nitrogen values were 0. 79" oand O. 59".

‘ Not determined.

respectively.

0.05 level (Student's t test). This suggests that some
non-UA-N initially present in the termite was
expended during UA biosynthesis. From the data in
Table 1, it can be calculated that 262°; of the initial
endogenous non-UA-N was expended over the 15
month period. Also during this period, the
contribution of UA-N to the total nitrogen increased
from 531;, to almost 70‘."o (Table 1).

Assays at the commencement and after 9 months
captivity revealed that the Nz-fixing activity of R.
flavipes was 0.43 and 0.10 iig nitrogen fixed per g fresh
wt per day. respectively.

The total carbon content of termites during the first
6 months of captivity of population 1 revealed a
decrease from 5057?0 to 46.57", of the dry weight.
This was remarkable as uric acid carbon (UA-C)
increased from 1.12" of the dry wt to 17. 16° during

this period (calculated from data in Table 1).
Apparently, 40.52;, of the initial endogenous non-UA-
C, i.e. [(50.57- 1.12) - (46.57- 17.16)]/(50.57— 1.12)
was also expended during UA biosynthesis.

While not influencing the foregoing calculations. it
may be noted that the average dry wt of R. flavipes
workers varied between 0.63 and 1.10 mg in a manner
that bore no apparent correlation with changes in UA,
total nitrogen. or total carbon (Table 1). The variation
in the average dry wt during the 15 month period may
have been due to reproductive activity in the
population (ESENTHER, 1977).

Accumulation of UA by termites was not only
detectable by enzymatic analyses, but could also be
inferred by visual inspection of live specimens.
Accumulation of white, chalky material in fat body
tissue accompanied the increase in UA (Fig. 4).

40

respectively. w hereas total

Dissection and careful removal of abdominal fat body
tissue revealed that it contained from 93"., to 95"., of
the insects’ total UA stores (two experiments; pooled
tissue from five insects). Little or no UA was
associated with excised guts.

Analysis of excreta and nest wood of R. flavipes

R. flavipes faecal pellets selected from nest wood
galleries were about 0.5 x 0.25 mm in size and
possessed slight longitudinal grooves presumably
formed by the termites‘ rectal papillae during pellet
dehydration and expression. Their colour ranged from
light tan to moderate brown, with no evidence of a ‘salt
and pepper' appearance (COCHRAN. 1973). These
pellets contained only trace amounts of UA.
accounting for less than 0.04“D of the pellet dry weight.
The total nitrogen content of the pellets was 0.52",,.
Freshly-voided faeces from R. flavipes workers
contained only 0.18‘3o UA. despite the fact that the
termites used for the experiment contained 15°, UA by
weight.

Dried milled nest wood contained 0.36‘.’0
nitrogen, but no detectable UA.

total

Distribution of uric acid in various termite species

Specimens of R. flavipes collected in Mississippi. as
well as those of five other termite species representing
the Rhinotermitidae and Kalotermitidae. all
possessed UA in amounts ranging from about 1-1420
of the insects' dry weight and accounting for 4—38‘,‘.’, of
the insects‘ total nitrogen (Table 2). The presence of
UA in all specimens was also verified by TLC with
solvent systems A and C. No UA was detected in
faecal pellets of P. sirnplicicornis or M. hubbardi.

Figure 4 .

41

Deposition of uric acid in fat body tissue of _R_. flav
workers. Specimen A was freshly-collected and contained
1.5% uric acid (w/w) . whereas specimen B was true a
population held in captivity for 15 months and contained
45.5% (w/w) . Note the intense white colcuration of fat body
tissue of specimen 8 due to uric acid deposition.

 

Uric acid in termites 25

although such pellets contained about 0.79;, total
nitrogen (Table 2, footnote ‘).

Assay for uricase activity in R. flavipes tissues

No uricase activity was detected in homogenates of
whole worker termites. bodies with guts removed, or
excised guts. However. commercial uricase (hog liver,
type V), when added to termite homogenates, retained
full activity even after admixtures were subjected to
dialysis (see Materials and Methods). These data
indicated that termite homogenates did not contain
substances inhibitory to uricase activity. Control
homogenates prepared from mouse liver readily
degraded UA at a rate of 7.26 nmoles per min per mg
protein. This rate was comparable to that reported by
SCHNEIDER and Hocanoom (1952).

Other compounds present in R. flavipes extracts

TLC of R. flavipes extracts revealed consistently
three compounds in addition to UA that were
detectable by U.V. illumination of plates. The
compounds. designated T-l, T-2, and T-3, were
observed in freshly-collected specimens of R. flavipes
as well as in specimens containing high amounts of
UA. Identification of the compounds was made by
comparing them with standards in terms of their: R,
values; visual appearance under short i. U.V. light; and
reactivity with DPC spray reagent. T-l was tentatively
identified as kynurenic acid after TLC in two solvent
systems. The R, value in solvent system A was 0.44
while that of standard kynurenic acid was 0.46. Both
compounds fluoresced blue-green under U.V. light
and were DPC negative. T-2 was tentatively identified
as inosine after TLC in six solvent systems. The R
value in solvent system A was 0.37, compared wit
0.39 for commercial inosine. It appeared dark grey
under U.V. light and was DPC positive. T-3 gave an R,
value of 0.27 in solvent system A, migrating just ahead
of UA (R, = 0.16). T-3 fluoresced blue-violet under
U.V. light and was DPC negative. The identity of T-3
is not known. Kynurenic acid and T-3 were associated
primarily with extracted guts of R. flavipes, whereas
inosine was observed only in extracts of whole termites
or bodies from which guts had been removed.

DISCUSSION

Results presented in this paper showed conclusively
that xylophagous termites formed UA in amounts
ranging from 1-45"., of their dry body weight and
accounting for 3—69‘}(, of their total nitrogen (Tables 1
and 2). While these values seem surprisingly high, they
are not unreasonable: MULLINS and COCHRAN (1976)
found UA to constitute 8—31"/o of the dry weight and
4—72‘3; of the total nitrogen of cockroach species they
examined. Moreover, as with cockroaches and certain
other insects (COCHRAN, 1975), the principle site of
UA deposition in termites was fat body tissue. This
tissue may also be the site of UA synthesis in termites.
In light of significant amounts of UA in R. flavipes, it
was not unusual to also detect inosine in extracts of
this species. Inosine is a precursor for UA synthesis by
either the nucleicolytic or uricotelic pathway
(COCHRAN, 1975). Inosine, like UA, was not found in
the gut. Kynurenic acid, however, was associated

43

primarily with guts of R. flavipes. Kynurenic acid may
be derived from a side reaction in the biosynthesis of
ommochromes from tryptophan (LINZEN, 1974) and,
therefore, not be directly involved in termite purine
metabolism. Kynurenic acid has also been found in
American cockroach excreta (MULLINS and COG-IRAN,
1973).

Despite appreciable levels of UA in the termites
themselves. faecal analyses revealed that R. flavipes, P.
simplicicornis and M. hubbardi’ did not void significant
amounts of this purine. 0f the small amount of
nitrogen in faecal pellets (approx. 0.633. w/w) less than
3°; of that was attributable to UA-N. Thus. these
species, as well as most cockroaches (COCHRAN, 1973.
1976). dispose of their UA by the mechanism of
‘storage excretion‘ (MADDRELL, 1971; COCIIRAN.
1975) and constitute exceptions to the general notion
that terrestrial insects should void UA (NEEDHAM.
1938).

Freshly-collected R. flavipes contained about 2°,
UA which increased to 452;, during 15 months of
captivity. It is not yet known what factors cause UA
accumulation in termites or whether accumulation of
UA to such an extent also occurs in nature. However,
it seems likely that termites accumulate UA because
their intake of dietary nitrogen exceeds that needed for
biosynthesis. This was clearly shown for American
cockroaches fed diets high in casein (MULLINS and
COCHRAN, 1975a). What was surprising with R.
flavipes, however, was the fact that appreciable
amounts of endogenous non-UA-N and non-UA-C.
equivalent to 26 and 41‘}; of the termites' initial total
nitrogen and total carbon, respectively, were expended
during UA biosynthesis over the 15 month period.
This does not necessarily mean that such nitrogen and
carbon were incorporated into the UA molecule,
although that seems likely. Nevertheless, it must be
remembered that whole termites were used for total
nitrogen and total carbon assay. Therefore, the origin
of such endogenous nitrogen and carbon expended
could have been termite tissue, or gut contents, or
both.

Expenditure of endogenous nitrogen could not
alone have accounted for the increase in UA and total
nitrogen. however; obviously exogenous nitrogen must
have also beenexploited.Twolikely sourcesof exogenous
nitrogen were nest wood and termite protoplasm (i.e.
via cannibalism). For example, the average dry weight
of termites at commencement and after 9 months
captivity was 0.81 mg. whereas UA levels were 1.7 and
32.4‘,’,,, respectively (Table 1). The net increase iii
UA-N was therefore 0.083 mg per termite at an
expenditure of 0.046 mg endogenous nitrogen. If all
the latter were incorporated into UA and the balance
(i.e. 0.037 mg nitrogen) was derived solely from nest
wood (which contained 0.36%, nitrogen by weight).
wood consumption would have had to have been
about 10 mg wood per termite per 9 months or 0.036
mg wood per termite per day. HAVERTY (1976)
reported a rate for wood consumption by R. flavipes
(0.083 mg per termite per day) which could easily have
supplied the necessary nitrogen, even assuming that
the assimilation of nitrogen has an efficiency of 50°,
(HUNGATE. 1941; 1944). A similar argument may be
made for cannibalism. Considering the average dry
weight of R. flavipes to be about 1 mg. one can

26 C. J. POTRIxus AND J. A. BREZNAR

calculate from the data in Table I that the increase in
UA-N during 15 months would require 0.12 mg
exogenous nitrogen or roughly 0.01 mg exogenous
nitrogen per termite per month. Assuming the total
nitrogen content of workers to be 10",, by weight,
cannibalism would have had to have occurred at a rate
of at least one worker devoured per ten survivors per
month to furnish the required nitrogen for UA
synthesis. This level is. in fact, significantly less than
that actually observed for Zootermopsis angusticollis
maintained on artificial diets (HENDEE, 1935),
although the extent of cannibalism in natural termite
colonies has not been thoroughly documented
(LEEAGE and NUTTING. 1978). N 2 Fixation by the gut
microbiota might also contribute nitrogen for UA
synthesis. although N , fixation rates of R. flavipes are
relatively low compared to some other termite species
(BREZNAK. 1975) and were low for the present
specimens. Clearly. however. several exogenous
nitrogen sources, alone or together. could furnish
enough nitrogen to easily account for the observed
accumulation of UA during laboratory captivity.
Our findings are consistent with the hypothesis of
LEACH and GRANOVSKY (1938) in regard to termites‘
ability to form UA. However the value, if any, of UA
accumulation to termites remains to be determined. In
what must be considered a classic study. MULLINS and
COCHRAN (1975a, b) showed that UA accumulated by
cockroaches (Periplaneta americana) on a positive
nitrogen balance diet was readily mobilized on
negative nitrogen balance diets, and a considerable
portion of UA-N was used then for obthecal
production by females. Interestingly. UA was not
voided even during periods of extensive mobilization;
NH3 constituted the main nitrogenous compound
released. Although P. americana is reported to possess
uricase (CORDERO and LUDWIG, I963), DONNELLAN
and KILBY (1967) suggested that bacterial symbionts
present in the fat body of this species also function in
uric acid mobilization. It is not known whether
cockroach gut bacteria also carry out uricolysis. but a
prominent bacterial flora is indeed present in the
alimentary tract of P. americana (BIONELL, 1977:
BRACKE et al., 1978). By contrast. we could not detect
uricase activity in R. flavipes tissues and have not yet
thoroughly examined fat body tissue for the presence
of endosymbiotic bacteria. However. significant
populations of uricolytic bacteria were present in
hindguts (PorRIRUS and BREZNAR, manuscript in
preparation). It is tempting to speculate that termites
store UA as a reserve material but. as suggested by
LEACH and GRANOVSRY (1938). require the activity of
uricolytic gut bacteria in order to recover carbon,
nitrogen and/or energy from the stored purine. If so,
the assertion of MOORE (1969), that UA is merely a
‘waste product‘ of termites. may require revision.

Note

Recent. ultrasensitive assays for uricase by the
method of FRIEDMAN and MERRIL (1973) also failed to
reveal this activity in R. flavipes Malpighian tubules or
fat body tissue. Control assays, employing single
Malpighian tubules from newly emerged adults of
Drosoplii/a melaiiogastcr. yielded typical positive
results (FRIEDMAN and JOHNSON, 1977) whether such
tubules were mixed with termite preparations or not.

Acknowledgements—We are indebted to Dr. DONALD G.
COcHRAN for advice. encouragement and for comments on
the manuscript prior to submission. We thank Dr. RICHARD
W. MERRITT for help in photographing Fig. 4 and Drs.
MICHAEL 1. HAVERTY and SYDNEY L. TAMM for supplying
some of the termite Specimens. Mass spectrometry of
tetraethyl UA was performed at the mass spectrometry
facility of Michigan State University. a service supported by
National Institutes of Health Grant RR00480. This
investigation was supported in part by the Michigan
Agricultural Experiment Station of Michigan State
University; by National Science Foundation Grants BMS 75-
05850 and PCM 77-05732, and by an NSF Predoctoral
Fellowship awarded to C. J. PorRiitus.

REFERENCES

BIONELL D. E. (1977) Some observations on the distribution
of gut flora in the American cockroach. Periplaneta
americana. J. Invert. Path. 29. 338-343.

Burn 1. W.. CRL'DEV D. L. and MARKOVETZ A. J. (1978)
Effect of metronidazole on the intestinal microflora of the
American cockroach. Periplaneta americana L.

Antmiicroh. Ag. A. 13, 115—120.

BREZNAIt J. A. (1975) Symbiotic relationships between
termites and their intestinal microbiota. Symp. Soc. exp.
Biol. 29. 559—580.

BRLZNAR .1. A., BRILL W. J.. MERriNs J. W. and Gwen H. C.
(1973) Nitrogen fixation in termites. Nature. Land. 244.
577-580.

COCHRAN D. G. (1973) Comparative analysis ofexcreta from
twenty cockroach species. C amp. Biochem. Physiol. 46A.
409-419.

COCHRAN D. G. (1975) Excretion in insects. 1n Insect
Biochemistry and Function (Ed. by C ANDY D. J. and KILBY
B. A.). pp. 178-281. Chapman 8: Hall. London.

COCHRAN D. G. (1976) Excreta analysis on additional
cockroach species and the house cricket. C amp. Biochem.
Physiol. 53A. 79—81.

CORDERO S. M. and LL'DwiG D. (1963) Purification and
activities of purine enzymes from various tissues of the
American cockroach. J. N. Y. Ent. Soc. 71. 66-73.

DixerIN S.. BEROMANN F. and CHAIMOVIrz M. (1956)
Studies on uric acid and related compounds—11. Paper
chromatography of substituted xanthines and uric acids. J.
biol. Chem. 221. 239-251

DONNELLAN .1. F. and Klik‘i B A. (1967) Uric acid
metabolism by symbiotic b‘JL‘IL‘llJ from the fat body of
Periplaneta americana. C omp. Biochem. Physiol. 22.
235-252. '

ESENTHER G. R. (1977) Fluctuation of survival. growth. and
dilTerentiation in biweekly samples of Reticulitermes
flavipes workers. Material org. 12. 49—58.

FRIEDMAN T. B. and JOHNSON D. H. (1977) Temporal control
of urate oxidase activity in Drosophila: evidence of an
autonomous timer in Malpighian tubules. Science 197,
477-479.

FRIEDMAN T. B. and MERRIL C. R. (1973) A micro-
radiochemical assay for urate oxidase. Ana/it. Biochem.
55. 292-296.

HAvERrv M. 1. (1976) Termites. Pest Control 14, 12-17. 46.
47. 48.

HENDEE E. C. (1935) The role of fungi in the diet of the
common damp-wood termite. Zootermopsis angusticollis.
Hilgardia 9. 499-525.

HERBERT D.. PHIPPS P. J. and STRANGE R. E. (19711Chemical
analysis of microbial cells. In Methods in Microbiologi-
(Ed. by NORRIS R. and RIBBONS D. W.). vol. SB. Academic
Press. New York.

HUNGATE R. E. (1941) Experiments on the nitrogen economy
of termites. Ann. Ent. Soc. Am. 34. 467—489.

44

Uric acid in termites 27

Ht'NOATE R. E. (1944) Termite growth and nitrogen
utilization in laboratory cultures. Tex. Acad. Sci. Proc.
Transac. 27. 1-7.

ISMAIL A. A. A. and DARIN T. A. ( 1975) Gas chromatography
and characterization of tetraethyl derivatives of uric acid.
J. Chromat. 110. 182-186.

JOHNSON M. J. (1941 ) Isolation and properties of a pure yeast
polypeptidase. J. biol. Chem. 137, 575—586.

JUCCI C. (1921) Sulla presenza di depositi uratici nei tessuto
adipose dei termitidi. Atti. Accad. naz. Lincei Memorie 30,
213—215.

LAFAGE J. P. and NUTTINO W. L. (1978) Nutrient dynamics of
termites. In Production Ecologi' of Ants and Termites (Ed.
by BRIAN M. V.). International Biological Programme Vol.
13. pp. 165—244. Cambridge University Press, London.

LEACH J. G. and GRANovSIcv A. A. (1938) Nitrogen in the
nutrition of termites. Science 87, 66-67.

LINztx B. t 1974) The tryptophan _. ommochrome pathway
in insects. Adv. Insect Physiol. 10. 117-246.

MADDRELL S. H. P. (1971) The mechanisms of insect
excretory systems. Adv. Insect Physiol. 8, 199—331.

MOORE B. P. (1969) Biochemical Studies in termites. In

Blip/HQI iii Termites (Ed. by KRISH\A K and WEESNER F. l

M.). vol. 1. pp. 407-432. Academic Press. New York.
ML‘LLINS D. E. and COCHRAN D. G. (1973) Tryptophan
metabolite excretion by the American cockroach. Comp.
Biochem. Physiol. “B, 549-555.
ML'LLiNs D. E. and COCHRAN D. G. (19753) Nitrogen

metabolism in the American cockroach—1. An
examination of positive nitrogen balance with respect to
uric acid stores. C amp. Biochem. Physiol. 50A, 489-500.

MULLINS D. E. and COCHRAN D. G. (1975b) Nitrogen
metabolism in the American cockroach—11. An
examination of negative nitrogen balance with respeCI to
.mobilization of uric acid stores. C amp. Biochem. Physiol.
50A. 501-510.

MULLINS D. E. and COCHRAN D. G. (1976) A comparative
study of nitrogen excretion in twenty-three cockroach
species. Comp. Biochem. Physiol. 53A. 393-399.

NEEDHAM J. (1938) Contributions of chemical physiology to
the problem of reversibility in evolution. Biol. Rev. 13.
225-251.

PATAItI G. (1967) Thin-layer chromatography of nucleic acid
bases. nucleosides, nucleotides and related compounds.
111. Separation of complex mixtures on cellulose layers. J.
Chromat. 29, 126—132.

POTRIRUS C. J. and BREzNAIt J. A. (1977) Nitrogen-fixing
Enterobacter agglomerans isolated from guts of wood-
eating termites. Appl. envir. Microbiol. 33. 392-399.

.SCHNEIDER W. C. and HOGENROOM G. H. (1952) Intracellular

distribution of enzymes. IX. Certain purine-metabolizing
enzymes. J. biol. Chem. 195, 161-166.

SCHULTZ J. E. and BREZNAR J. A. (1978) Heterotrophic
bacteria present in hindguts of wood-eating termites
[Reticulitermes flavipes (Kollar)]. Appl. envir. Microbiol.
35, 930-936.

45

ARTICLE II

URICPCID-mmm
INQTI‘SWW

[mm mm txomil

C.J. Pctrikus and J.A. Breznak

mprinted fran Appl. Divircn. Microbiol. 19:117-124 (1980)

46

47

Amen AND ENVIRONHEN‘TAL Micnoaiotocv. July 1960. p. 117-124

”9.2240/80/07-0117/08302111/0

Vol. 40. No. I

Uric Acid-Degrading Bacteria in Guts of Termites
[Reticulitermes flavipes (Kollar)]T
C. J. POTRIKUS AND JOHN A. BREZNAK'

Department of Micmbiologv and Public Health, Michigan State University, East Lansing. Michigan 48824

Uricolytic bacteria were present in guts of Reticulitermes flavipes in popula-
tions up to 6 x 10‘ cells per gut. Of 82 strains isolated under strict anaerobic
conditions, most were group N Streptococcus sp., Bacteroides termitidis, and
Citrobacter sp. All isolates used uric acid (UA) as an energy source anaerobically.
but not aerobically, and NH: was the major nitrogenous product of uricolysis.
However, none of the isolates had an absolute requirement for UA. Utilization of
heterocyclic compounds other than UA was limited. Fresh termite gut contents
also degraded UA anaerobically, as measured by “C02 evolution from [2-"C]UA.
The magnitude of anaerobic uricolysis [0.67 pmol of UA catabolized/ (gut x h)]
was entirely consistent with the population density of uricolytic bacteria in situ.
Uricolytic gut bacteria may convert UA in aim to products usable by termites for
carbon, nitrogen, energy. 0r all three. This possibility is consistent with the fact
that R. flavipes termites form UA, but they do not void the purine in excreta

despite the lack of uricase in their tissues.

 

Uric acid (UA) is a major nitrogenous excre-
tory product of many terrestrial insects (6, 7),
and it is well suited to this purpose. Because of
its poor solubility in water, UA can be defecated
as a nontoxic solid. Water loss during excretion
is thereby minimized (6, 7). However, UA should
not be regarded as merely a nitrogenous waste
product of insects. For example, Periplaneta
americana cockroaches store UA internally and
use it as a metabolic reserve when placed on
nitrogen-deficient diets (24, 25). UA nitrogen
mobilized under such conditions is used in part
for oothecal production by females (25). Utili-
zation of DA apparently occurs in some other
insects as well (7).

Intuitively, one might expect that xylopha-
gous termites, which thrive on nitrogen-poor
diets, would have evolved efficient means of
conserving combined nitrogen. One strategy to
accomplish this was envisioned by Leach and
Granovsky (19). These workers hypothesized
that UA, elaborated into the termite gut via the
Malpighian tubules, is degraded by the hindgut
microbiota to a form of nitrogen reusable by the
insects. A corollary to their hypothesis is that
carbon atoms of UA are also reused. Leach and
Granovsky performed no experiments to test
their provocative notion, although the abun-
dance of microbes in the termite hindgut (4, 5)
makes their suggestion plausible.

A major interest in our laboratory is the role
of gut organisms in termite nutrition. We there-

‘lJournal article no. 9312 from the Michigan Agricultural
Experiment Station.

fore sought to critically test the UA-recycling
hypothesis. In a previous paper (28), we reported
that termites form UA, but they do not void the
purine despite the lack of uricase in termite
tissues. We now present data to reconcile this
apparent anomaly and which support Leach and
Granovsky‘s hypothesis. Herein we document
the presence of uricolytic bacteria in guts of
Reticulitermes flavipes termites. The present
paper deals with the isolation, identification. and
nutrition of the bacteria, as well as estimation of
their population levels in situ. An accompanying
paper (29) concerns the anaerobic metabolism
of UA by pure cultures.

MATERIALS AND METHODS

Insects. R. flavipes (Kollar) termites were collected
in Janesville, Wis, Dansville. Mich, and Spring Arbor.
Mich., and were maintained in the laboratory as pre-
viously described (31). UA content of termites was
determined by enzymatic assay after extraction from
tissue with Li2COJ (28).

P. americana L. cockroaches were obtained from
the Pesticide Research Center of Michigan State Uni-
versity. They were maintained on a commercial dog
food diet and were fed water ad lib.

Media and culture techniques. Strict anaerobic
techniques were used for preparation of liquid and
solid media (15, 17, 31) unless indicated otherwise.

The composition of culture media is reported as
percentage (wt/vol). unless indicated otherwise. Dou-
ble-layer agar plates were used to isolate uricolytic
bacteria and contained 0.1 and 1.0% UA in the bottom
and top layers, respectively (1). Isolation media were
designated SUA. BHIU. and TYU. SUA was the sup-
plemented uric acid agar of Barnes and Impey (1) with

117

48

118 POTRIKUS AND BREZNAK

the following exceptions: 0.09% beef extract (Difco
Laboratories, Detroit. Mich.) plus 0.15% peptone
(Difco) replaced Lab.Lemco, and 0.1%. liver concen.
trete (Wilson Diagnostics) replaced liver extract.
BHIU medium was supplemented brain-heart infusion
(Bl-II) medium (15) containing UA as indicated above.
TYU agar medium contained: tryptone (Difco), 1.0%;
yeast extract (Difco), 0.1%; salts solution (15) minus
NaHCO 4. 4% (vol/vol); cysteine - HCl, 0.05%: resazurin,
10"? and UA as indicated above. All solid media
contained 1.5'11r agar.

TYU broth medium consisted of TY basal medium
into which NaOH-solubilized UA was incorporated.
TY basal was similar to TYU agar (above). but lacked
UA. agar. resazurin. and cysteine - HCl. For addition to
sterile TY broth. UA was solubilized as a 2% solution
in 0.5 N NaOH, filter sterilized (filter type GS; 0.22-
pm pore size; Millipore Corp., Bedford. Mass), and
added at a final concentration of 0.1%. A predeter-
mined amount of sterile 0.5 N HCl was then added to
neutralize the medium before inoculation. TYFU
broth medium was identical to TYU, but also con-
tained filter-sterilized fructose at a final concentration
of 0.049. In general. TY-based broth media were pre-
pared in air by heat sterilization and rapid cooling of
the TY basal portion. followed soon thereafter by
additions as described above and then replacement of
the head space atmosphere with 02-free gas. However,
for culture of bacteroides the TY basal portion was
prereduced and anaerobically sterilized (15). Culture
vessels were then transferred into a vinyl anaerobic
glove box (Coy Manufacturing Co., Ann Arbor. Mich.)
containing an Orfree atmosphere of Nz/Hg (90/10.
vol/vol) wherein appropriate additions to the medium
were made. The head space atmosphere of culture
vessels was replaced with Nz/co, (95/5, vol/vol) after
inoculation.

Semi-D basal medium. used to conduct fermenta-
tion balances with growing cells, contained: Trypticase
(BBL Microbiology Systems. Cockeysville. Md.), 1.0%;
salts solution (15) minus NaHCO;. 4% (vol/vol); MnClz -
41130, 0.00592; adenine, thymine, guanine. cytosine,
and uracil. 5 x 10% each; riboflavin, calcium panto-
thenate. nicotinic acid, pyridoxine- HCl. and folic acid,
10‘“; each; thiamine-HCl, 10”‘%; and biotin, 107%. A
chemically defined medium. D medium, was identical
to semi-D medium but lacked Trypticase.

Small volumes (e.g., 10 ml) of broth media were
usually contained in 18-mm anaerobe tubes (Bellco
Glass, Inc.. Vineland, NJ.) equipped with rubber stop-
pers. Mass culture of cells was done by using aspirator
bottles of various sizes and employing O-z-free N2 in
the head space. Aerobic broth culture was as previ-
ously described (27).

Isolation of uricolytic bacteria. Removal of ter-
mite guts and isolation of bacteria was done as de-
scribed previously (27, 31). except that gut homoge-
nates were prepared in TY broth and treated in a
microblender (Eberbach) for 60 s before dilution and
plating. Isolation plates were incubated in an anaero-
bic glove box and were supplemented with CO; (31).
After 7 days, presumptive uricolytic isolates were ran-
domly picked from among those colonies surrounded
by a clear zone in the otherwise opaque medium (1).
Such colonies were easily recognizable, even against a

Area. Envraon. MICROBIOL.

background of loo-fold excess nonuricolytic colonies.
Isolates were considered to be pure cultures after three
successive passages on streak plates; microscopic ob-
servations. including Gram stains of cells, verified this
conclusion.

Control experiments used the following inocula: sur-
face washes of intact termites; and homogenates pre-
pared from degutted termite bodies.

Characterization of isolates. General character-
ization of isolates was done as previously described
(27, 31). Acidic fermentation products were deter-
mined after anaerobic growth of isolates in Bl'II con-
taining 0.59": glucm (Bl-IIG). Media for tests with
Bacteroides isolates were prepared as described by
Dowel] and Hawkins (11). Spore formation was rou-
tinely evaluated by observing wet mounts or malachite
green-stained preparations of cells by phase-contrast
or bright-field microscopy. respectively. For Bacte-
roides isolates. spore formation was assessed while
cells were growing on chopped meat agar slants for 3
weeks (15). In addition, the heat resistance of Bacte-
roides in starch broth was tested (15).

The moles percent guanine plus cytosine (G+C)
content of deoxyribonucleic acid (DNA) of Bacte-
roides termitidis was determined by buoyant density
analysis. Cells were grown in TY broth supplemented
with 1% glucose, 0.024% DL-threonine, and [methyl-
aI-I]thyrnidine (10 “Ci/ml) to radioactively label the
DNA. DNA was partially purified by the method of
Marmur (21). except that only two deproteinization
treatments were used, and these were followed by a
single precipitation step with isopropanol. The buoy-
ant density of such preparations in CsCl gradients was
determined as described by Schildkraut et al. (30).
Samples were centrifuged at 20°C in a Beckman model
50T i rotor at 40,000 rpm for 60 h. Centrifuge tubes
were then punctured from the bottom, and fractions
(10 drops each) were collected. Fifty microliters of
each fraction was applied to trichloroacetic acid-im-
pregnated filter paper squares, which were then
washed. dried, and analyzed for radioactivity (12). The
refractive index of fractions containing peak radioac-
tivity was measured by using an Abbe 3-L refractom-
eter (Bausch & Lomb, Inc., Rochester, N.Y.)., and
from this the moles percent G+C in DNA was calcu-
lated (30). DNA from Escherichia coli B (p - 1.710;
30). prepared as described above but labeled by growth
of cells with [methyl-"Clthymidine (0.1 uCi/ml cul-
ture). was used as an internal standard.

Nutrition and growth studies. Most nutrition
studies were done with the growtholimiting TY basal
medium (above) to which additions were made. Het-
erocyclic compounds to be added were solubilized as
described above for UA. Exceptions were inosine, pu-
rine, and caffeine (which were prepared as 2‘? stock
solutions in water). and xanthine (which was prepared
as a 1% stock solution in 0.5 N NaOI-I). All test
substrates were filter sterilized (see above). Where
necessary, pH adjustment was made with HCl (above)
before inoculation. Some nutritional studies were done
with medium D.

Growth of cells was measured turbidirnetrically at
660 nm by using a Spectronic 20 calorimeter (Bausch
& Lomb). Absorbance readings were converted to cell
numbers by means of standard curves relating the

49

V01. 40. 1980

reading to direct microscopic counts (Streptococcus)
or viable cell counts (Bacteroides and Citrobacter).
Unless otherwise stated, cultures were incubated at
30°C (Streptococcus) or 37°C (Bacteroides and Citro-
bacter).

Assay for UA, NH;. and urea in spent growth media
was done after removal of cells by centrifugation at
12.000 x g for 10 min. UA was assayed spectrophoto-
metrically at 292 nm by using hog liver uricase (uric
acid diagnostic kit no. 292-UV; Sigma Chemical Co.,
St. Louis. Mo.). Urea was assayed by detennining
urease-dependent NH; production with phenol nitro-
prusside reagent (urea nitrogen diagnostic kit no. 640;
Sigma). NH; was determined by using the urea assay
reagents without urease.

Anaerobic metabolism of DA by minced ter-
mite guts. Guts were removed as described above
from 100 R. flavipes workers and pooled into a small
watch glass containing 500 pl of 0.1 M potassium
phosphate buffer (pH 7.0). Guts were then minced
with a scalpel. and transferred to a 13-mm test tube,
and the suspension was made up to 1,500 pl with
buffer. The suspension was then subject to agitation
in a Vortex blender at high speed for 20 5. Five hundred
microliters of such preparations was used per reaction
mixture. Individual reaction mixtures (620 pl) con-
tained: potassium phosphate buffer (pH 7.0), 52 pmol;
dithiothreitol, 1 pmol; [2-"C]UA, 2 to 3 nmol; sodium
formate (optional), 1 mol; and 30 to 40 minced gut
equivalents.

Reactions were performed under 90% N2/ 10% H2 in
5—ml serum vials fitted with stoppers. Suspended from
each stopper was a center well assembly (Kontes,
Vineland, N. J.; catalog no. K-882320). Reactions were
initiated by addition of UA and were terminated after
3 h by injection of 150 pl of 5 N HCl. Phenethylamine
(200 pl) was then injected into the center well and
allowed to absorb l‘CO; for one additional hour, after
which time radioactivity was estimated.

Radioactivity measurements. Radioactivity
measurements and quench corrections were made as
described previously (32).

Microscopy of insect fat body tissue. Fat body
tissue was dissected into a primary fixative of 2‘7:
glutaraldehyde in 0.1 M potassium phosphate buffer
(pH 7.4) and then held at 4°C for 2 h. After primary
fixation. tissue was washed three times in glutaralde-
hydevfree phosphate buffer and postfixed in phosphate
buffer containing 1‘2 Os0.. Specimens were then de-
hydrated with ethanol, cleared with propylene oxide,
and embedded in Epon (20). Sections (1 to 3 pm thick)
were cut with a diamond knife and were viewed di-
rectly by phase-contrast microsc0py or by bright-field
microscopy after staining with buffered azure ll (18).

Other bacterial strains. For comparative pur-
poses and as positive or negative controls, various
known bacterial strains were used. Termite gut het.
erotrophic bacteria previously isolated on nonselective
medium (31) were part of this laboratory‘s culture
collection. Proteus mirabilis. Citrobacter freundii,
Staphylococcus aureus, and E. coli B were obtained
from the culture collection of the Department of Mi-
crobiology and Public Health. Michigan State Univer-
sity. Streptococcus lactis ATCC 19435. S. cremoris
ATCC 19257, S. faecalis ATCC 19433, and S. faecium

URICOLYTIC GUT BACTERIA OI" TERMITES

119

ATCC 19434 were obtained from the American Type
Culture Collection, Rockville. Md. S. lactis C2 was
obtained from L. L. McKay, Department of Food
Science and Nutrition, University of Minnesota, St.
Paul.

Chemicals. Heterocyclic substrates were obtained
from Sigma Chemical Co. and were of the highest
purity available. [2-"C]UA was obtained from Amer-
sham Corp., Arlington Heights. Ill. All other chemicals
were of reagent grade and were obtained from various
commercial sources.

RESULTS

Isolation and identification of uricolytic
bacteria. Over a period of 3 years, uricolytic
bacteria were consistently isolated from guts of
R. flavipes, regardless of the termites’ origin or
length of captivity (Table 1). Population levels
ranged from <102 to 6.1 x 10‘ cells/gut, with a
grand mean of 3.5 t 1.7 X 10‘ (excluding the
data from experiments 7 and 8). We do not know
why so few uricolytic bacteria were detected in
experiments 7 and 8 (Table 1). Control experi-
ments (Table 1, footnote (1) verified that urico-
lytic isolates were of gut origin. Light microscopy
failed to reveal bacteria within termite fat body
tissue cells. However, typical mycetocytes (i.e.,
specialized host tissue cells containing endosym-
biotic bacteria) (8) were readily observed in fat
body tissue of P. americana cockroaches used
as a control.

The UA content of termites ranged from 0.4
to 31.1% of the insects’ dry body weight (Table
1). As documented previously (28), UA stores of
R. flavipes generally increase during laboratory
captivity, yet no clear correlation existed be-
tween the UA content of termites and the num-
ber of uricolytic gut bacteria.

Of 117 strains initially isolated, 35 lost the
uricolytic phenotype upon subculture from pri-
mary isolation plates, and they were not exam-
ined further. Of the 82 strains that retained the
phenotype as a fairly stable characteristic, most
were identified as group N Streptococcus sp., B.
termitidis, and Citrobacter sp. One strain of
group D Streptococcus was obtained.

General physiological and biochemical char-
acteristics of the group N streptococcal isolates
are presented in Table 2. Lactate was the major
acidic fermentation product (Table 2). suggest-
ing that cells were homolactic. Detailed studies
with one isolate (strain UAD-l) verified this
notion. When growing anaerobically in semi-D
medium containing excess (i.e., 1‘?) glucose.
strain UAD-l formed (millimoles/ 100 mmol of
glucose fermented): lactate, 162.5; acetate. 21.1;
ethanol, 5.2; formate, 5.0; and C02. 7.0. Similarly,
products from 1% fructose were (millimoles/IOO
mmol of fructose fermented): lactate, 168.4; ace-

 

 

 

 

120 POTRIKUS AND BREZNAK APPL. Envmos. MICROBIOL.
TABLE 1. Uricolytic bacteria present in guts of R. flavipes
Tsrmites‘ identification of isolates
. No. of

firm length UA con gm :3- 'uolates “MN P 0'3” Citro-

no. . . of cap« tent (Q . exam- 8. ter-

Onsin‘ 0m collected am of dry um/sut‘ M 5‘"? fig; “1“" mmdu.
(mo)' wt) p.
€448 0448

I J June 1976 1 ND' 3.8 X 10‘ 20 18 0 2 0

2 4 ND 3.9 X 10‘ 23 13 l 9 0

3 10 ND 2.8 x 10‘ 15'

4 J June 1977 o 1.7 1.8 x 10‘ 20’

5 4 15.1 4.5 X 10‘ 8 0 0 0 8

6 5 24.6 1.5 X 10‘ 3 0 0 0 3

7 10.5 31.1 <10” 0

8 J June 1978 0 2.0 <102 0

9 8 18.0 3.2 X 10‘ 10 0 0 10 0
10 J June 1979 0 3.3 0.6 X 10‘ 4 1 0 0 3
11 D June 1979 0 4.0 5.6 X 10‘ 6 6 0 0 0
12 2 0.4 6.1 X 10‘ 3 3 0 O 0
13 D August 1979 o 1.7 4.9 x 10‘ 5 5 o o o

 

‘ All termites were workers, except for experiment 6 (brachypterous larvae).

°J, Jancsville, Wis; D, Dansville, Mich.

' Zero months refers to termites used within 24 h of collection.
‘ Mean value of two to six replicates made with SUA plates. No uricolytic bacteria were detected in surface
washes of intact termites or in homogenates of degutted termite bodies.

' ND, Not determined.

’ Isolates lost uricolytic ability upon subculture and were not studied further.

tate, 22.9; ethanol, 13.8; formate, 16.6; and C02,
9.1. However, the overall properties of group N
isolates did not correspond exactly to those of a
known species (9, 13). Uricolytic strains failed to
produce NHa from arginine (distinguishing them
from S. lactis), yet showed growth at 40°C, at
pH 9.2, and in the presence of 0.3% methylene
blue (distinguishing them from S. cremoris)
(Table 2). Preliminary examination of lactate
dehydrogenase activity of strain UAD-l sug-
gested similarities to that of S. faecium, a group
D Streptococcus (13; E. Garvie, personal com-
munication). The mol% G+C in the DNA of
strain UAD-l was 36.6 (E. Garvie, personal com-
munication), a value in the middle of the range
found for the genus Streptococcus. Based on
these results, we deferred species assignment of
the group N isolates.

B. termitidis isolates were acidogenic rods 0.5
X 3.1 m in size. Their properties corresponded
closely to those of known B. termitidis (14, 33),
although the present isolates did not produce
formate as a fermentation product and formed
small amounts of acid from lactose in a delayed
(5 days) reaction. In addition, all strains pro-
duced acid from mannitol, mannose, rhamnose,
and trehalose, but not from melezitose. The
mol% G+C in the DNA was 35.6 (strain UAD-
50).

Citrobacter isolates were indole positive, sim-
ilar to strains previously isolated on nonselective
media (31). These were not assigned to a species.

Nutritional studies. Nutritional studies on
the isolates were done to confirm UA utilization;
to examine the role of UA as a carbon, nitrogen,
and energy source for growth; and to evaluate
their nutritional diversity with respect to hetero-
cyclic compounds in general. UA increased the
anaerobic cell yield of all isolates when the pu-
rine was incorporated into an otherwise growth-
limiting medium (i.e., TY medium). These data
suggested that energy was derived from anaer-
obic uricolysis by the isolates. Results with three
representative strains are presented in Table 3.
Preliminary experiments with Streptococcus
UAD-l revealed that small amounts (0.04%) of
fructose markedly stimulated UA consumption
and cell yields. Consequently, this sugar was
frequently added to TY medium for evaluating
the utilization of heterocyclic compounds by all
strains. Interestingly, diauxic growth of strain
UAD-l (i.e., a primary logarithmic growth phase
separated from a small secondary growth phase
by 10 to 20 h) resulted when 0.04% glucose was
included with UA. Under these conditions UA
was consumed only slightly (i.e., 10% of initial)
and only during the secondary phase, after all
the glucose was depleted. Diauxic growth was
not observed when fructose was included with
UA.

Anaerobic cell yields of streptococci and bac-
teroides were at least twice as great with UA as
with identical media lacking this purine (Table
3). In addition, NH; production was 79 to 1009?

51

V01. 40, 1980

TABLE 2. Characteristics of uricolytic group N
Streptococcus isolates from R. flavipes

 

 

Test or substrate Reaction'
Gram reaction Variable
Morphology Ovoid; 0.6 to
0.7 pm in
diam
Motility -
Acidic fermentation products Lf (as)
Final pH (BHIG medium) 4.5
Relation to 02 Facultative
Catalase —
Nitrate to nitrite '-
Growth in:
BHI + 4.0% NaCl +
BHI + 6.5% NaCl —
0.1% methylene blue milk 4»
0.3% methylene blue milk 4»
Growth in BHI at:
10°C +
40°C +
45°C -t+)‘
pH 9.2 +
pH 9.6 +
Heat tolerance“ —
Hydrolysis of:
Esculin +
Hippurate -
Arginine -
Hemolysis a
Precipitin reaction with group N +‘
antiserum

 

" Symbols: +, Positive reaction; -. negative. Acidic
fermentation products: L, lactic; f, formic; a, acetic; s,
succinic. Upper-case and lower-case letters refer to
major and minor amounts, respectively (15). Paren-
theses indicate results with some strains.

° Two of 46 strains showed growth at 45°C.

‘ 60°C. 30 min.

‘ Only 12 strains tested.

of theoretical (based on UA consumed).

Although Citrobacter isolates readily cleared
UA-containing plates, their cell yields were not
greatly stimulated by UA in broth media (Table
3). Under the latter conditions, -only small
amounts of DA were consumed by Citrobacter
strains, yet such UA consumption was accom-
panied by NH; production. Interestingly, fruc-
tose suppressed UA utilization and NH; produc-
tion by all Citrobacter strains.

UA was not used as an energy source aerobi-
cally by streptococci or citrobacters. This con-
clusion was based on growth yield determina-
tions as well as enzymatic assay of UA consump-
tion.

The ability of isolates to use other heterocyclic
compounds anaerobically was fairly limited (Ta-
ble 3). Streptococcus UAD-l appeared to cata-
bolize allantoin and allantoic acid only slightly,
as indicated by a small increase in cell yield and

URICOLYTIC GUT BACTERIA OI" TERMITES

121

NH; production (Table 3), as well as formation
of small amounts of urea (Table 3, footnote d).
Increased growth of this strain on inosine and
on fructose plus guanosine apparently resulted
from utilization of the ribose portion of the mol-
ecules only, since: (i) the final pH of the media
(pH 5.0 to 5.1) was more acidic than that of UA-
containing media (pH 7.2 to 7.3); (ii) little or no
NH3 or urea was formed; (iii) the purine portion
of the ribosides accumulated in (and in the case
of guanosine alone, crystallized out of) the me-
dium and was readily identified by thin-layer
chromatography, and (iv) hypoxanthine and
guanine alone were not used although ribose
was. No other purines or pyrimidines tested were
used by strain UAD-l.

B. termitidis UAD-50 appeared to use inosine
and hypoxanthine; utilization of xanthosine was
questionable. Interestingly, xanthine and gua-
nosine inhibited growth of strain UAD-50. Citro-
bacter UAD~25 showed slightly more nutritional
versatility than either strain UAD-l or UAD-50.
Utilization of guanosine, hypoxanthine, xan-
thine, and xanthosine was obvious from the in-
crease in NH; production. Inosine utilization was
questionable: although cell yields were in-
creased, there was no increase in NH; produc-
tion.

By using a chemically defined medium (me-
dium D), we found that none of the representa-
tive strains used UA as sole carbon, nitrogen,
and energy source for anaerobic growth. Strep-
tococcus UAD-l would not use 0.1% UA as sole
nitrogen or sole energy source with 0.4% fructose
or 1.0% Trypticase, respectively, but would use
UA as an additional energy source with 0.04%
fructose plus 1.0% Trypticase. Two B. termitidis
strains tested behaved identically to Streptococ~
cus UADoI, except that one strain (UAD-55)
used UA as sole nitrogen source in the presence
of 0.4% fructose.

Comparative tests with other bacterial
strains. Reference strains of group N strepto-
cocci (i.e., S. lactis ATCC 19435 and S. cremoris
ATCC 19257) did not use UA when growing
anaerobically in TYU or TYFU broth media,
nor did these strains form clear zones on TYU
agar plates. Tests were also done on S. lactis C2,
as well as 54 group N streptococci previously
isolated from guts of R. flavipes (31) and which
included S. lactis, S. cremoris, and strains of
unknown species. None used UA anaerobically
as judged by lack of clear zone formation on UA-
containing plates. Of Citrobacter strains isolated
previously (31), only indole-positive biotypes
cleared UA-containing plates. Previously iso-
lated Bacteroides strains (31) were not tested
for uricolytic ability, although the characteris-

52

 

 

 

 

122 POTRIKUS AND BREZNAK APPL. Enviaon. MICROBIOL.
TAIL: 3. Anaerobic utilization of uric acid and other heterocyclic compounds by growing cells“
Streptococcus UAD~1 B. termitidis UAD~5O Citrobacter DAD-25
Added‘ to TV heal me- . . .
. Yield‘ 4 Yield Yield
drum NH, produced NH; produced NH; produced
‘2":47" (pmol/ml) ‘3“:47" (pmol/ml) ‘fufé'f" (pmol/ml)

No addition 2.0 0.0 0.4 1.2 1.3 7.3

Uric acid 4.2 7.3 (2.3) 1.4 17.3 (4.9) 1.4 12.8 (1.4)

F 5.8 0.2 1.3 0.7 1.9 4.5

I" + uric acid 11.5 23.5 (6.1) 2.6 23.3 (6.1) 2.4 5.1 (0.3)

Allantoin 2.3 2.3 0.4 1.3 1.1 7.9

F + allantoin 6.4 2.5 1.4 0.9 ND' ND

Allanwic acid 1.8 0.4 0.4 1.0 1.0 8.3

F + allantoic acid 6.6 3.6 1.4 0.8 ND ND

Inosine 5.8 0.3 2.1 7.8 3.0 7.0

F + inosine 6.3 0.3 2.4 5.8 ND ND

Guanosine —’ 0.2 0 ND 3.3 18.7

F + guanosine 7.1 0.3 0 ND ND ND

Hypoxanthine 1.9 0.0 2.3 8.9 3.0 16.2

F + hypoxanthine 6.1 0.2 2 7 9.1 ND ND

Xanthine 1.3 0.3 0 ND 1.6 15.7

F + xanthine 4.7 0.4 0 ND ND ND

Xanthosine 1.5 0.0 0.5 ND 2.5 25.3

F + xanthosine 5.5 0.0 1.3 4.3 ND ND

Riboae 3.3 0.0 0.9 1.2 1.3 6.0

F + ribose 7.5 0.0 ND ND ND ND

 

‘ The following compounds were not used by growing cells: adenine, guanine. purine. caffeine, thymine, cytosine, uracil. and

orotic acid.

° Final concentration of all heterocyclic compounds was 0.1% except for guanine (0.05%). Fructose (F) and ribose were used
at a concentration of 0.04 and 0.05%, respectively. Incubation was under N./co. (95/5, vol/vol). Initial pH of media was 7.0 1

0.2.

' Maximum yields achieved between 24 and 48 b (UAD-l) or 36 and 96 h (UAD-5O and UAD-25). A yield of 0 designates <5

x 10‘ cells/ml.

‘ Trace amounts (0.05 to 0.7 pmol/ml) of urea were formed by strain UAD-l from allantoin and allantoic acid only. Urea was
not formed from uric acid by strains UADo50 and DAD-25. Decomposition of any of the heterocyclics to yield NH; or urea did
not occur in uninoculsted media. Values in parentheses indicate micromoles of uric acid consumed per milliliter during growth;

initial concentration of uric acid was 6.1 pmol/ml.
' ND, Not determined.

’Ceflyieldsweremtdeterminsdbecauaegmnmeaystahformedintbemedium

tics of such strains differed significantly from
those of B. termitidis isolates obtained in the
present study.

Anaerobic metabolism of UA by minced
termite guts. We attempted to detect anaerobic
metabolism of UA in fresh termite gut contents.
Since 60% or more of UA carbon is evolved as
CO; by isolates (29), "002 evolution from [2-
“CJUA was used as an estimate of uricolysis by
minced guts. Results (Table 4) showed that such
preparations degraded a significant amount of
UA under strict anaerobic conditions. Little or
no stimulation of uricolysis was achieved by
including formate in reaction mixtures.

The magnitude of uricolysis by minced guts
was consistent with the density of uricolytic
bacteria in guts (see Discussion).

DISCUSSION

Results presented herein show that uricolytic
bacteria are present in guts of R. flavipes, gen-
erally in populations of 3.5 :I: 1.7 X 10‘ cells/gut.
This density is roughly 10% that of total culti-
vable heterotrophs (31). Nevertheless, we be-
lieve such populations to be numerically signifi-

cant, based on arguments developed previously
(31) regarding the tiny size of R. flavipes guts.
Results are consistent with Leach and Granov-
sky’s hypothesis (19), and also with our previous
findings (28) that R. flavipes termites form UA
but do not void the purine despite their lack of
uricase. Presumably UA is catabolized by bac-
teria when it enters the gut, and therefore it is
not found in the feces. However, the dynamics
of DA mobilization and bacterial uricolysis in
situ, as well as the nutritional control of these
processes, are important points which remain to
be defined.

Uricolytic isolates were most likely derived
from the hindgut, although a small portion of
midgut usually remains attached to extracted
hindguts (5, 31). Previous studies (31) showed
that populations of heterotrophs obtained from
midgut homogenates were only 1/no that of the
present uricolytic isolates. However, Malpighian
tubules of R. flavipes empty at the precise junc-
tion of mid- and hindgut, i.e., at the enteric valve
(26). Thus, some uricolytic bacteria could con-
ceivably colonize the midgut, near the enteric
valve, yet still have ready access to UA.

53

V01. 40, 1980

TABLE 4. Anaerobic metabolism of [2-"C1UA by
minced guts of R. flavipes workers"

 

 

“CO, evolved pmol of sub-
Added‘ to reac- —_ strate "C
tion mixture evolved as
(nmol) dpm pmol "CO,/gut
equivalent
[2-“C]UA (2.5) 7.601 67.8 2.06
[2-"C]UA (2.5) + 7,707 68.7 2.08
sodium formate
(1.000)
Control 0 0 O

 

' Termites collected in Spring Arbor, Mich. and
held captive for 7 months before assay.

‘Specific activity of [2-"C]UA was 50 nCi/nmol.
Individual reaction mixtures contained 33 gut equiva-
lents. Incubation was at 25°C for 3 h. Control mixture
(lacking formate) was terminated with HCl immedi-
ately after UA addition.

Three major groups of uricolytic bacteria were
associated with R. flavipes: group N Streptococ-
cus sp., B. termitidis, and Citrobacter sp. Their
generic affiliation was the same as that of major
heterotrophs isolated previously on nonselective
medium (31). Barnes and Impey (1) found a
greater phenotypic diversity of uricolytic bacte-
ria isolated from chicken ceca, although strep-
tococci and bacteroides were among those iso-
lated. Fecal streptococci of human and sheep
origin have also been shown to degrade UA (23).
Our isolation of uricolytic B. termitidis parallels
that of Sebald (33), who isolated this species
(referred to by her as Sphaerophorus siccus var.
termitidis) from gut contents of Reticulitermes
lucifugus (14). By contrast, Donnellan and Kilby
(10) reported the isolation of an aerobic, urico-
lytic vibroid bacterium from fat body tissue of
American cockroaches. Uricase mediated the
first step of uricolysis by the vibroid (10). As
reported herein, we have never observed bacte-
ria of any type in fat body tissue of R. flavipes.
Moreover, ultrasensitive assays for uricase failed
to reveal such activity in fat body tissue (28).

Estimates of anaerobic UA metabolism by
minced gut preparations suggested an activity of
about 2 pmol of UA catabolized/ (gut equivalent
x 3 h) (Table 4), which is equivalent to 0.67
pmol/(gut x h). Considering that 10” cells of
Streptococcus UAD-l or B. termitidis DAD-50
can degrade approximately 20 pmol of UA dur-
ing a 2-h incubation (see Fig. 2 and 3 and Table
2, 3, and 5 of reference 29). it can be calculated
that the in situ density of such bacteria would
have to be about 0.67 x 10‘ cells/gut to account
for the activity of minced gut preparations. The
latter value compares extremely well with re-
sults of enumeration studies (Table 1), indicat-
ing that the uricolytic activity of fresh gut con-

URICOLYTIC GUT BACTERIA OF TERMITES

123

tents can be completely accounted for by isolates
obtained in pure culture.

Uricolytic isolates displayed limited nutri-
tional versatility with respect to heterocyclic
compounds other than UA (Table 3). This was
especially true for Streptococcus UAD-l. Such
metabolic specialization of the isolates undoubt-
edly relates to the ability of their termite host to
form UA (28). The ultimate benefit of uricolytic
bacteria to R. flavipes, however, can only be
inferred at the present time. Acetate, a major
carbonaceous product of uricolysis by the iso-
lates (29), is also a major product of cellulolysis
in the termite gut (4) and it is a source of carbon
and energy for the insect (2, 4, 22). On the other
hand, the status of NH3, the major nitrogenous
product of uricolysis, is less clear. Ammonia
nitrogen could cycle back to the insects directly,
or indirectly after assimilation into microbial
protoplasm or other microbial metabolites as
suggested by Leach and Granovsky (l9). Inter-
estingly, such a cycle appears to occur between
the UA-forming marine flatworm Convoluta
roscoffensis (which also lacks uricase) and its
intracellular, uricolytic algal symbiont Platy-
monas convolutae (3, 16). In this case, NHJ
formed during uricolysis by P. convolutae is
apparently used by the alga for synthesis of
amino acids. some of which are exported to the
flatworm host. Whatever may be the exact flow
of UA nitrogen in R. flavipes, it seems clear that
uricolysis by gut bacteria could be extremely
important to nitrogen conservation within the
termite colony.

ACKNOWLEDGMENTS

We are grateful to Clifford Sporn for excellent technical
assistance and to Donald G. Cochran for helpful discussions.
We thank E. I. Garvie for determining the moles percent G+C
content in the DNA of strain UAD-l and for amaying lactate
dehydrogenase in that strain.

This work was supported in part by the Michigan Agricul-
tural Experiment Station of Michigan State University and
by National Science Foundation grants BMS 75—05850 and
PCM 77-05732. C.J.P. held a National Science Foundation
predoctoral fellowship during the major portion of this study.

LITERATURE CITED

1. Barnes, E. M., and C. S. Impey. 1974. The occurrence
and properties of uric acid decomposing anaerobic bac-
teria in the avian caecum. J. Appl. Bacteriol. 37:393-
409.

2. Blomquist, G. J., R. W. Howard, and C. A. McDaniel.
1979. Biosynthesis of the cuticular hydrocarbons of the
termite Zootermopsis angusticolis: (Hagen). Incorpo-
ration of propionate into dimethylalltanes. Insect Bio-
chem. 9:371-374.

3. Boyle, J. E., and D. C. Smith. l975. Biochemical inter-
actions between the symbionts of Convoluta rosco/fen-
sis. Proc. R. Soc. Lond. Ser. B 189:121-135.

4. Breznak, J. A. 1975. Symbiotic relationships between
termites and their intestinal microbiota. p. 559-580. In
D. H. Jennings and D. L. Lee (ed.). Symbiosis (Society

54

124 POTRIKUS AND BREZNAK

5.

for Experimental Biology Symp. Ser. no. 29). Cambridge
University Press. Cambridge.

Breznak. J. A., and H. S. Pankrata. 1977. In situ
morphology of the gut microbiota of woodoeating ter-
mites [Reticulitermes flavipes (Kollar) and Copto-
termes formosanus Shiraki]. Appl. Environ. Microbiol.
33:406-426.

6 Chapman. R. I". 1975. The insects: structure and function.

7.

ll.

12.

13.

I4.

15.

16.

17.

American Elsevier Publishing Co., Inc., New York.
Cochran, D. G. 1975. Excretion in insects, p. 178-281. In
. D. J. Candy and B. A. Kilby (ed.). Imect biochemistry
and function. Chapman and Hall. London.

. Cochran. D. G., D. E. Mullins, and K. J. Mullins. 1979.

Cytological changes in the fat body of the American
cockroach. Periplaneta americana, in relation to die-
tary nitrogen levels. Ann. Entomol. Soc. Am. 72:197-
205

. 6...... it 11., and H. w. Seeley, Jr. 1974. Genus 1.

Streptococcus Rosenbach 1884, 22. In R. E. Buchanan
and N. E. Gibbom (ed.). Bergey's manual of determi-
native bscten'ology. 8th ed. The Williams 4: Wilkins
Co., Baltimore, Md.

. Donnellan, J. F., and B. A. Iiilby. 1967. Uric acid

metabolism by symbiotic bacteria from the fat body of
Periplaneta americana. Comp. Biochem. Physiol. 22:
235-252.

Dowel), V. R., Jr., and T. M. Hawkins. 1968. Labora-
tory methods in anaerobic bacteriology. U.S. Depart-
ment of Health. Education. and Welfare, Atlanta. Ga.

Furano. A. V. 1971. A very rapid method for washing
large numbers of precipitates of proteins and nucleic
acids. Anal. Biochem. 43:639-640.

Garvie, E. I. 1978. Streptococcus rafftnolactis Orla-Jen-
sen and Hansen. a group N streptococcus found in raw
milk. Int. J. Syst. Bacteriol. 28:190-193.

Holdeman, L. V., and W. E. C. Moore. 1974. Genus l.
Bacteroides. Castellani and Chalmers 1919. 959. In R.
E. Buchanan and N. E. Gibbons (ed.). Bergey's manual
of determinative bacteriology, 8th ed. The Williams It
Wilkins Co., Baltimore, Md.

Holdeman, L. V., and W. E. C. Moore (ed.). 1975.
Anaerobe laboratory manual. Virginia Polytechnic In-
stitute and State University. Blacksburg.

Holligan, P. M., and G. W. Gooday. 1975. Symbiosis in
Convoluta roscof/ensis, p. 205—227. In D. H. Jennings
and D. L. Lee (ed.). Symbiosis (Society for Experimeno
tal Biology Syrnp. Ser. no 29). Cambridge University
Press. Cambridge.

Hungate, R. E. 1969. A roll tube method for cultivation
of strict anaerobes, p. 117-132. In J. R. Norris and D.
W. Ribbons (ed.). Methods in microbiology, vol. 38.
Academic Press. Inc., New York.

. Jeon, K. W. 1965. Simple method for staining and pre-

21.

25.

31.

. Sebald. M. 1962. Etude sur les bactéries

ArrL. ENVIRON. MICROBIOL.

serving epoxy resin embedded animal thsue sections for
light microscopy. Life Sci. 4:1839—1841.

Leach, J. G., and A. A. Granovsky. 1938. Nitrogen in
the nutrition of termites. Science 87:66-67.

. Luft. J. H. 1961. Improvements in epoxy resin embedding

methods. J. Biophys. Biochem Cytol. 9:409—414.

Mai-mur, J. 1961. A procedure for the isolation of deoxy-
ribonucleic acid from micro-organisms. J. Mol. Biol. 3:
208-218.

. Mauldin. J. K., N. M. Rich, and D. W. Cook. 1978.

Amino acid synthesis from I‘C-acetate by normally and
abnormally faunated termites. Coptotermes formo.
sanus. Insect Biochem. 8:105-109.

. Mead. G. C. 1974. Anaerobic utilization of uric acid by

some group D streptococci. J. Gen. Microbiol. 82:421-
423.

. Mullins, D. E., and D. G. Cochran. 1975. Nitrogen

metabolism in the American cockroach. I. An exami-
nation of positive nitrogen balance with respect to uric
acid stores. Comp. Biochem. Physiol. 60A:489-500.

Mullins, D. E., and D. G. Cochran. 1975. Nitrogen
metabolism in the American cockroach. II. An exami-
nation of negative nitrogen balance with respect to
mobilization of uric acid stores. Comp. Biochem. Phys-
iol. 50Az501-510.

. Noirot, C., and C. Noirot-Timothée. 1969. The digestive

system, p. 49-88. In K. Krishna and F. M. Weesner
(ed.). Biology of termites. vol. 1. Academic Press. Inc.
New York.

. Potrikus, c. J., and J. A. Breznak. 1977. Nitrogen-

fixing Enterobacter agglomerans isolated from guts of
woodeating termites. Appl. Environ. Microbiol. 33:
392—399.

. Potrikus, C. J., and J. A. Breznak. 1980. Uric acid in

wood-eating termites. Insect Biochem. 10:19-27.

. Potrikus, C. J., and J. A. Breznak. 1980. Anaerobic

degradation of uric acid by gut bacteria of termites.
Appl. Environ. Microbiol. 40:125-132.

. Schildkraut, C. I... J. Marmur, and P. Doty. 1962.

Determination of the base composition of deoxyribo-
nucleic acid from its buoyant density in CsCl. J. Mol.
Biol. 4:430—443.

Schultz, J. E., and J. A. Breznak. 1978. Heterotrophic
bacteria present in hindguts of wood-eating termites
[Reticulitermes flavipes (Kollar)]. Appl. Environ. Mi-
crobiol. 35:930-936.

. Schultz, J. E., and J. A. Breznak. 1979. Crosafeeding

of lactate between Streptococcus lactis and Bacteroides
sp. isolated from termite hindguts. Appl. Environ. Mi-
crobiol. 37:1206-1210.

anaerobies
gram-negatives asporulees. lmprimerie Barneoud S. A.,
Laval.

ARI‘ICLE III

ANAEKBIC WOW (I URIC PCID

BYGJ'I‘BPCI‘ERIACFW

C.J. Potrikus and J.A. Breznak

Reprinted fruit Appl. Ehviron. Microbiol. iQdZS-IBZ (1980)

SS

13.-I

!

Arman AND ENVIRONMENTAL Micsostowcr. July 1980. p. 125-132

(1199-2240/m/07-0125/08302fll/0

Vol. 40. No 1

Anaerobic Degradation of Uric Acid by Gut Bacteria of

TermitesT

C. J. POTRIKUS AND JOHN A. BREZNAK'

Department of Microbiology and Public Health, Michigan State University, East Lansing, Michigan 48824

A study was done of anaerobic degradation of uric acid (UA) by representative
strains of uricolytic bacteria isolated from guts of Reticulitermes flavipes termites.
Streptococcus strain UADol degraded UA incompletely, secreting a fluorescent
compound into the medium, unless formate (or a formicogenic compound) was
present as a cosubstrate. Formate functioned as a reductant, and its oxidation to
CO; by formate dehydrogenase provided 2H‘ 4» 2e‘ needed to drive uricolysis to
completion. Uricolysis by Streptococcus UAD-l thus corresponded to the follow-
ing equation: lUA + lformate -o 4C0: + lacetate + 4NH3. Urea did not appear
to be an intermediate in C02 and NH; formation during uricolysis by strain UAD-
l. Forrnate dehydrogenase and uricolytic activities of strain UAD-l were inducible
by growth of cells on UA. Bacteroides termitidis strain UAD-50 degraded UA as
follows: lUA -e 3.5 C0: + 0.75acetate + {NH-i. Exogenous formate was neither
required for nor stimulatory to uricolysis by strain DAD-50. Studies of UA
catabolism by Citrobacter strains were limited, because only small amounts of
UA were metabolized by cells in liquid medium. Uricolytic activity of such
bacteria in situ could be important to the carbon, nitrogen, and energy economy

of R. flavipes.

 

It is fair to say that much is known about
anaerobic degradation of uric acid (UA), but in
an extremely limited number of bacteria—pri-
marily Clostrt'dt'um acidiurt'ct' and C. cylindros-
porum (23). However, anaerobic uricolysis is a
property fairly widespread among nonspore-
forming bacteria (2, 3, 12, 19, 25), and animal
intestinal tracts appear to be a rich source of
uricolytic organisms (3, 9, IO, 12, 13, 19). Unfor-
tunately, little is known concerning the major
products of uricolysis by pure cultures of gut
bacteria, let alone the stoichiometry of UA ca-
tabolism.

In a companion paper (17) we reported on the
isolation, identification, and nutritional charac-
teristics of uricolytic bacteria from the guts of
Reticulitermes flavipes termites. In this paper
we report on anaerobic degradation of UA by
representative isolates from the termite gut: a
group N Streptococcus; Bacteroides termitidis;
and a Citrobacter sp. The purpose of this study
was twofold: (i) to increase our understanding of
the role(s) of gut bacteria in termite nutrition;
and (ii) to contribute to our knowledge of anaer-
obic purine degradation in general.

MATERIALS AND METHODS

Bacteria, media, and culture techniques. Bac-
teria, media, and culture techniques are described in
the companion paper (17).

tJournal Article no. 9358 from the Michigan Agricultural
Experiment Station.

Analysis of fennentations by growing cells.
Fermentations with growing cells were performed by
using a fermentation train (14). UA was assayed spec-
trophotometrically at 292 nm by using hog liver uricase
(uric acid diagnostic kit no. 292-UV; Sigma Chemical
Co., St. Louis, Mo.). Samples for UA assay at the end
of the fermentation were removed before acidification
and clarification (14) of the medium. Cells were re-
moved from such samples by centrifugation at 12,000
x g for 10 min, and the supernatant fluid was used for
UA assay. Most other substrates and products were
assayed as previously described (15, 18). Fructose was
assayed enzymatically as described by Bernt and Berg-
meyer (4). Urea was assayed by determining urease-
dependent N H; production with phenol nitroprusside
reagent (urea nitrogen diagnostic kit no. 640; Sigma).
NH; was routinely determined by using the urea assay
reagents (above) without urease. However. for fermen-
tation balances, NH; was first collected by microdif-
fusion before assay (7). Oxarnate was assayed as de-
scribed by Valentine and Wolfe (22). Glycine was
estimated by visually observing the intensity of glycine
spots after one-dimensional thin-layer chromatogra-
phy (TLC) and staining with ninhydrin. Uninoculated
medium served as a control. For TLC of glycine, 0.1-
mm-thick cellulose MN 300 layers (Brinkmann Instru-
ments Inc., Woodbury. NY.) were used with a solvent
system of isopropanol-formic acid-water (40:2:10).
Heterocyclic compounds were routinely separated by
using one-dimensional TLC with cellulose MN 300
layers (above) and the following solvent systems: It-
butanol-methanol-water-NI-LOH (60:20:20:1) and
isopropanoI-water (7:3) (16). Compounds were visual-
ized by illumination with shortwave ultraviolet (UV)
light or by spraying plates with Erlich reagent (5).

Oxidation/reduction (O/ R) indexes were calculated

I25

5'7

126 POTRIKUS AND BREZNAK

by the method of Neish (14) for sugar fermentations
and by the method of Whiteley and Douglas (25) for
fermentations involving UA.

Preparation of cell suspensions. cell extracts.
and reagents for anaerobic metabolism studies.
Cells were grown anaerobically to late logarithmic or
early stationary phase in TYU medium. They were
then harvested by centrifugation at 6,000 x g for 10
min. Cell pellets were washed once with, and sus-
pended in, 0.2 M potassium phosphate buffer to a final
density of about 10" cells/ml. The pH of the phos-
phate buffer was 7.0, except when cells were being
suspended for UA metabolism (pH 6.4) or for assay of
formate dehydrogenase (FDH) (pH 8.0).

Cell extracts were prepared by first adding lysozyme
(0.5 mg/ml, final concentration) to cell suspensions
and allowing them to incubate for 30 min at 30°C.
Suspensions were then brought to about 4°C and
exposed to three bursts (l5 s/burst at 30-s intervals)
from the probe of a Bronwill Biosonik II sonic oscil-
lator operating at full power. Preparations were then
centrifuged at 18,000 x g for 10 min, and the resulting
supernatant fluids were used as cell-free extracts.

All manipulations of cells and extracts were done in
the cold (2 to 5°C) under N2, with minimum exposure
of the preparations to air. Heat-stable reagents (e.g.,
buffers) were boiled and quickly cooled under Oz-fiee
N2 before use. Suspensions or heat-labile solutions
were held in serum-stoppered bottles. the gas phase of
which was vacuum evacuated and flushed at least four
times before being sealed.

Metabolism of cell suspensions. Reaction were
carried out in double-side-arm Warburg vessels by
using conventional manometric techniques (21). This
facilitated measurement of gaseous as well as soluble
products. Products were sometimes determined from
the combined contents of several vessels containing
identical reaction mixtures. Individual reaction mix-
tures (3.0 ml) contained: potassium phosphate buffer,
305 pmol; dithiothreitol, 10 nmol; approximately 10"
cells; and a variable amount of substrate as indicated.
The pH of phosphate buffer was 7.0 unless UA was
the substrate, in which case it was pH 6.4. This assured
that when UA solution (0.1 M in 0.5 N NaOH) was
tipped into reaction mixtures (at a final concentration
of 20 umol/reaction mixture) the resulting pH would
be 7.0. A gas phase of N: was used throughout.

Substrates and products were generally assayed as
described for fermentations with growing cells (above).
“C02 was trapped in the center well of Warburg
vessels which contained NaOH-impregnated filter pa-
per. Filters were then pooled into a 2.5-mm test tube
containing 10 ml of water, 100 mg of nonradioactive
NaHC03, and a drop of cresol red indicator solution.
The tube was then sealed with a serum stopper,
through which 5 N HCl was injected to acidify the
mixture. I‘00.,» evolved in this way was then swept into
phenethylamine-methanol traps for radioactivity de-
terminations (18).

"C-labeled organic acids were separated by silicic
acid chromatography before determination of specific
radioactivity (14). "C-labeled ethanol was distilled
from clarified fermentation liquors at pH 7.5 (14);
radioactivity in such distillates was assumed to be due
solely to ethanol.

APPL. ENVIRON. Mrcaosron

O/R indexes were calculated as described for fer-
mentation with growing cells (above).

Radioactivity measurements. Radioactivity
measurements and quench corrections were made as
described previously (18).

Chromatography and spectrophotometry of
fluorescent compound FC-l. For TLC of F0), cell-
free fermentation liquors were concentrated by lyoph-
ilization and then redissolved to 1/ 10 original volume
with distilled water or 0.1 M potassium phosphate
buffer (pH 7.0). Samples of such material were applied
to TLC plates and developed with isopropanol-water
(above). F0! was subsequently visualized by illumi-
nation with UV light. When desired, F0] was scraped
from plates and eluted from the cellulose MN 300
support with phosphate buffer. Fluorescence spectra
were made by using an Aminco-Bowman spectropho-
tofluorimeter.

FDH. FDH activity of cell-free extracts was deter-
mined as described by Wagner and Andreesen (24).
except that dithiothreitol replaced dithioerythritol.
and benzyl viologen was used as the electron acceptor.

Chemicals. Heterocyclic substrates were obtained
from Sigma and were of the highest purity available.
Enzymes used in the assay of fructose were obtained
from Boehringer Mannheim Biochemicals. Indianap-
olis, Ind. Radioactive substrates and standards were
obtained from New England Nuclear Corp., Boston.
Mass. All other chemicals were of reagent grade and
were obtained from various commercial sources.

RESULTS

Anaerobic uricolysis by Streptococcus
UAD-l: growing cells. It was of fundamental
interest to determine the overall products of
uricolysis by growing cells. However, previous
experiments (17) indicated that limited dissimi-
lation of UA occurred unless fructose was in-
cluded in the medium. Accordingly, 0.04% fruc-
tose was included with UA in semi-D medium,
and a dual-substrate fermentation was carried
out. Products formed were corrected for those
formed in a separate fermentation of 0.04% fruc-
tose alone. Results (Table 1) did not permit
conclusions regarding the precise origin of car-
bonaceous products, but clearly revealed that
significantly greater quantities of C02 and ace-
tate were formed during UA fermentation. NI-l3
was the major nitrogenous product of UA fer-
mentation. Small amounts of lactate and ethanol
may have also been produced from UA. Surpris-
ing, however, was the absence of formate in
spent fermentation liquors. Thus, when cor-
rected for the formate produced during mono-
substrate (0.04% fructose) fermentation, a large
negative value was obtained (Table 1). This
indicated that, relative to 100 mmol of UA fer-
mented, 133.8 mmol of formate either was not
produced or was consumed. It should be noted
that batch culture fermentations in media con-
taining limited (i.e., 0.04%) fructose as sole fer-

Von 40. 1980

TABLE 1. Fermentation of UA and fructose by
growing cells of Streptococcus UAD- l "

 

mmol of product/ 100 mmol of:

 

Product
Fructose

 

(0.04%) 0‘
Lactate 73.7 37.8
Acetate 37.0 92.0
Ethanol 86.5 9.3
Formate 124 .4 -133.8
CO; 26.7 200.4
NH; 341.3
% C recovery 103.2 76.5
% N recovery 85.3
O/R index 1.0 2.6

 

“Growth medium was semi-D containing either
0.04% fructose alone or 0.04% fructose plus 0.1% UA.
Consumption of fructose was 99% in each fermenta-
tion, whereas consumption of UA was 35‘}.

°Tabulated data have been corrected for products
formed from fermentation of 0.04% fructose alone and
then normalized to 100 mmol UA fermented. The
following products were not detected: glycine, urea.
oxamate, H3, propionate, butyrate, succinate, fuma-
rate. pyruvate, acetoin, diacetyl, xanthine, hypoxan-
thine, thymine, uracil, orotic acid, allantoin, and allan-
toic acid.

mentable substrate were heterolactic, and for-
mate was a significant product (Table 1). This
was in marked contrast to the homolactic fer-
mentation pattern of strain UAD-l with excess
(1%) sugars (17). However, these data parallel
the observed shift to a heterolactic fermentation
when certain group N streptococci are grown at
low dilution rates in glucose-limited chemostat
cultures (20).

Carbon and nitrogen recoveries were generally
low in such dual-substrate fermentations, and
O/R indexes were at variance with an expected
value of 1.0, even if a more complex basal me-
dium such as TY was used (carbon recovery -
65%, nitrogen recovery - 66%; O/R index - 1.2).
Other plausible products of UA fermentation,
including reduction products such as xanthine
or hypoxanthine, were not detected (Table l,
footnote 5). However, during a search for the
latter by TLC of concentrated fermentation liq-
uors, a compound was observed which fluoresced
blue when illuminated by shortwave UV light.
This compound, termed FC-l, is discussed fur-
ther below, but production of F0] may account
for the relatively low carbon and nitrogen recov-
eries seen with dual-substrate fermentations.

Anaerobic uricolysis by Streptococcus
UAD-l: cell suspensions. In an effort to ob-
tain a simpler (i.e., monosubstrate) system for
studying UA fermentation by strain UAD-l,
UA-grown cell suspensions were incubated in
buffer with UA as the sole substrate. However,

ANAEROBIC URICOLYSIS BY TERMITE GUT BACTERIA

127

even cell suspensions required fructose for vig-
orous dissimilation of UA. By using C02 evolu-
tion as an index of UA dissimilation, little catab-
olism of the purine was apparent in the absence
of added fructose (Fig. 1). The greater the initial
amount of fructose included with UA, the
greater were the rate and extent of C02 evolu-
tion. When initial amounts of fructose and UA
were equirnolar (20 pmol each), the total C02
evolved was much greater than the sum of CO;
evolved from fructose and [IA alone. Little ad-
ditional stimulation of CO; evolution occurred if
fructose was increased to 40 pmol in the presence
of 20 pmol of UA. These data suggested that
some interaction between the substrates was
occurring.

To examine the relationship between fructose
and UA metabolism more closely, a “titration
experiment" was done to determine the amount
of fructose needed to effect complete dissimila-
tion of UA (Fig. 2). In the absence of added
fructose, about half of the initial UA was con-
sumed. This does not mean that the UA con-
sumed was completely catabolized, merely that
it disappeared as uricase-reactive material. In-
deed, complete catabolism of the consumed UA
seemed unlikely since little C02 and NH; were
evolved. However, addition of increasing initial
amounts of fructose resulted in more extensive

 

 

 

 

 

 

 

 

75 u u a
sotoaous
/.’.
s
/./
:3 t
s
a so I. s ”venous ‘
t2 0
O ./ so: o‘zous
8
- .
a ./
0 / ”orous
a . /. ‘ .
6 ° ./s/.
a 25 . /. ..
k / assoaoua
: ..-—--' '
s/./
./ 2004.
/ ’° ' .
. /!/: - A 3 ’0' :
'13,- v — v («609
o 0*. 0 01-— —
(.l 40 80 120
MINUTES

F10. 1. CO; evolution from fructose (F) and UA by
cell suspensions of Streptococcus UAD-I. The num-
bers indicate the amount (micromoles) of each sub
strate initially included in each Warburg reaction
vessel. Cells were grown in TYFU medium.

59

128 POTRIKUS AND BREZNAK

UA consumption and C02 and N Ii; evolution.
When the initial concentration of fructose was
equirnolar to UA, complete consumption of UA
occurred and theoretical amounts of NH; were
liberated. Under such conditions CO: evolution
was also copious.

Since complete UA dissimilation apparently
required at least equirnolar amounts of fructose,
a materials balance was performed in the dual-
substrate mode, but with [U-“C]fmctose. In this
way, the partitioning of UA and fructose carbon
into products could be evaluated. A control fer-
mentation of fructose alone was essentially
homolactic (Table 2, column 3). Moreover, the

specific activity of lactate carbon was nearly 4

identical to that of fructose carbon, indicating
that no dilution of 1‘C label occurred (Table 2,
footnotes a and b). For the dual-substrate fer-
mentation, the amount of fructose added was
roughly twice that of UA (Table 2, footnote a).
By determining the specific I‘C radioactivity in
each carbonaceous product, a materials balance
for each substrate was obtained. For ease of
evaluation, and since the amount of fructose
consumed was almost exactly twice that of DA
(Table 2, footnote 0), the data were transposed
to indicate products formed from 100 pmol of
fructose and from 50 pmol of UA. Results indi-
cated that C02, acetate, and NH3 were the major
products derived from fermentation of UA, al-
though small amounts of lactate and ethanol
were also formed. Inherent in this conclusion
was the assumption that any dilution of "C label

 

t1”

/ .

l h
o s 10 16 so
pmol Fructose Added

FIG. 2. Fructose-driven UA dissimilation by cell
suspensions of Streptococcus UAD-I. Each reaction
vessel initially contained 20 pmol of UA and a vari-
able amount of fructose as indicated. Reactions were
terminated after 130 min at 30°C, at which time gas
evolution had ceased.

incl (:02 or Ill-l3 Produced

 

 

APPL. ENVIRON. MICROBIOL.

TABLE 2. Fermentation of UA and I U- "leructose
by cell suspensions of Streptococcus UAD- I'

 

 

 

 

pmol of product per:
Product‘ 1“) pmol of
50 pmol 100 mol of
of UA * [U'Efm' [U-"C)fructoee

Lactate 4.2 149.8 145.6
Acetate 19.0 41.0 27.5
Ethanol 5.9 12.9 27.4
Formate 0.0 0.0 35.2
CD: 157.0 46.4 15.4
NH; 182.1

‘1: C recovery 96.8 99.5

‘5 N recovery 91.1

O/R index 1.0 1.2

 

 

 

" Individual reaction mixtures (3.0 ml) contained:
10” TYI-‘U-grown cells (equivalent to 19.5 mg. dry
weight); 51.0 pmol of [L'-“C]fructose (specific activity
- 5111 dpm/pmol of carbon) with or without 23.2 pmol
of UA; and other reagents listed in Materials and
Methods. Incubation was at 30°C for 2 h under N ,..
Consumption of substrates in the dual-substrate re.
action mixture was 47.0 pmol of [U-"leructose and
23.2 pmol of UA.

’ Specific activity (dpm per micromole of carbon) of
lactate produced in monosubstrate fermentation was
5,011. Specific activities of products in the dual-sub-
strate fermentation were: lactate, 4,971; acetate. 3,494;
ethanol, 3,500; and C0,». 1.165.

in soluble products, particularly acetate, oc-
curred in each carbon atom of the molecule (an
assumption borne out by further experiments
described below). Urea did not appear to be an
intermediate in NH; and 002 production. Urea
was never detected as a product of UA fermen-
tation, and UA-grown cells did not evolve co.
and NH3 from urea, although Proteus mirabilis
control cells did so readily.

Interestingly, products derived from fructose
in the dual-substrate fermentation (Table 2, col-
umn 2) were quantitatively similar to those
formed in the monosubstrate mode (Table 2,
column 3). An exception to this was the lack of
formate production in the dual-substrate fer-
mentation, accompanied by an increase in CO:
from fructose. This observation was similar to
that made with growing cells (Table l). The
total amount of C02 produced per 100 pmol of
fructose in the duaLsubstrate mode (46.4 pmol;
Table 2) was roughly equal to the sum of formate
and C02 produced in the monosubstrate mode.
This suggested that formate oxidation to CO,»
occurred during, and might be important to, UA
fermentation.

Formate-driven uricolysis by Streptococ-
cus UAD-l. To test the hypothesis that formate
oxidation was necessary for extensive uricolysis
by strain UAD-l, a titration experiment was

60

V01. 40. 1980

performed with formate by using a protocol sim-
ilar to that used with fructose. Results indicated
that formate, like fructose, also served to drive
uricolysis (Fig.3). Complete consumption of UA,
accompanied by liberation of theoretical quan-
tities of NH3, occurred when the initial quantity
of formate was equirnolar to that of UA. That
formate was oxidized to CO: during uricolysis
was shown with a companion reaction mixture
containing 20 nmol of UA plus 20 nmol of
[“C]formate (specific activity - 1,225 dpm/
pmol). Approximately 98.7% of the "C label was
recovered as "CO: at the end of the reaction.

Recognition of the importance of formate to
uricolysis prompted a materials balance for a
simplified dual-substrate fermentation of UA
(i.e., formate plus UA) (Table 3). For ease of
interpretation, the raw data were normalized to
100 pmol of UA fermented. Major products of
UA fermentation were C02, NH;, and acetate.
Almost all (92.4%) of the formate was recovered
as “C02. Thus it is clear that UA contributes
both the CH3 and COOH groups of acetate,
supporting the assumption made in evaluating
the data of Table 2 (above). However. since
radioactivity of acetate was not determined, it is
possible that small amounts of formate carbon
may have been used for acetogenesis. No evi-
dence of H2 production was observed.

These data revealed the link between fructose
metabolism and uricolysis by strain UAD-l (Fig.
2). Apparently, fructose metabolism was needed
to provide formate, whose oxidation served to

 

1 i
H

. N.‘ ‘C 1‘

 

 

 

g- ° .§
3» f
in ./ ms“
g..:/. \ i E
4. .1. f. 5%"

 

pmol Formete Added

F16. 3. Formate-driven UA dissimilation by cell
suspensions of Streptococcus UAD-l. Each reaction
vessel initially contained 20 pmol of UA and a vari-
able amount of sodium formate as indicated. Reac-
tions were terminated after 130 min at 30°C, at which
time gas evolution had ceased.

ANAEROBIC URICOLYSIS BY TERMITE GUT BACTERIA

129

drive uricolysis to completion. As anticipated,
other formicogenic compounds (e.g., glucose. ri-
bose, and pyruvate) also stimulated uricolysis
(data not shown). No stimulation was observed
with glycine.

FDH activity of Streptococcus UAD-l.
The importance of formate oxidation to urico-
lysis prompted a closer inspection of FDH activ-
ity in strain UAD-l. In experiments with cell
suspensions, I‘CO: evolution from ["C]formate
was taken as an index of FDH activity. Results
(Table 4) indicated that phenotypic expression
of FDH required that cells be grown in media
containing UA. However, even UA-grown cells
did not exhibit FDH unless an exogenous e‘
acceptor was present. The e' acceptor could be
a natural one (e.g., UA) or an artificial one (e.g.,
benzyl viologen) (Table 4). Methyl viologen and
methylene blue could also function as e’ accep-
tors for FDH, but they were not as effective as
benzyl viologen (data not shown).

Preliminary experiments revealed FDH activ-
ity in cell-free extracts of UAogrown cells. FDH
activity (121 nmol of formate oxidized x min“
x mg of protein") was apparently sensitive to
02: it could be completely abolished by exposure
of extracts to air for 10 min.

Anaerobic dissimilation of heterocych
compounds by cell suspensions of Strepto-
coccus UAD-l. Suspensions of UA-grown cells
of strain UAD-l evolved no significant CO.» and
NH; from the following substrates (equirnolar
fructose also included in reaction mixtures): xan-
thine, allantoin, allantoic acid, uracil, and orotic
acid.

Production of a fluorescent compound
FC-l by Streptococcus UAD-l. Growing cells,

TABLE 3. Fermentation of UA and ["leormate by
cell suspensions of Streptococcus UAD-I "

 

nmol of product per:

 

 

 

Produc" 125.6 pmol of 100.0 pmol of
["leormate UA
002 116.0 337-8
NHa 401.3
Acetate 88.5
‘ln C recovery 100.8
O/R index 0-9

 

" Reaction conditions similar to those described in
Table 2, footnote a. except that the substrate mixture
was ["leormate (specific activity - 10.019 dpm/pmol)
plus UA. Substrate consumption was: 19.6 of 20.0 pmol
of ["C]formate added and 15.6 of 20.6 pmol of UA
added.

" Specific activity (dpm/micromole) of "CO; was
2,557. Specific activity of acetate was not determined;
all acetate was assumed to be derived from UA.

61

130 POTRIKUS AND BREZNAK

as well as cell suspensions of strain UAD-l, often
secreted a fluorescent compound (F01) into the
medium. FC-l production was specific to UA
metabolism and was not detected during fer-
mentation of sugars alone. Moreover, control
experiments verified that FC-l was not a spon-
taneous decomposition product of UA. Condi-
tions which resulted in incomplete catabolism of
UA (i.e., fermentation of UA in the absence of a
cosubstrate) were those which also favored FC-
1 accumulation. For example, in the experiment
depicted by Fig. 2, supernatant fluids were ob-
tained from each reaction mixture and a portion
of each was applied to TLC plates. After ascend-
ing chromatography and visualization with UV
light, the F01 spot (R, - 0.15 to 0.20) was most
intense from the reaction mixture containing no
initial fructose. The intensity of the spot dimin-
ished from fluids which contained increasing
amounts of fructose until, with 20 pmol of initial
fructose, no FC-l spot was visually observed.
This was, of course, the reaction mixture result-
ing in complete uricolysis (Fig. 2). These data
suggested that FC-l was an intermediate in, or
by-product of, uricolysis by strain UADol.

When FC-l was eluted from TLC plates with
0.1 M potassium phosphate buffer (pH 7.0), a
fluorescence spectrum of the crude eluate re-
vealed excitation and emission maxima at 308
and 379 nm, respectively. A control spectrum (of
material eluted from a comparable position in a
TLC lane containing spent fructose fermenta-
tion liquor) revealed no significant fluorescence.
The identity of PO] is not yet known and awaits
purification of the compound.

Anaerobic uricolysis by B. termitidis. Cell

TABLE 4. Formate oxidation by cell suspensions of
Streptococcus UADJ

 

 

Growth Added to reaction :39;
substrate mixture (pmol) l l)
Fructose No additions 0.05
UA (20) 0.00
Benzyl viologen (10) 0.08
Fructose + No additions 0.04
UA UA (20) 6.48
Benzyl viologen‘ (10) 6.96
Benzyl viologen‘ (10) + 0.06

boiled cells

" Cells grown in TY basal medium containing 1.0%
fructose alone or 0.04% fructose plus 0.1% UA (i.e.,
TYFU medium).

‘ Reaction mixtures (3.0 ml) contained: 10" cells and
10 pmol of ["CJformate (specific activity - 96,493
dpm/pmol) plus phosphate buffer and dithiothreitol
as listed in Materials and Methods. Reaction condi-
tions: pH 7.0; 30°C; 60 min; gas phase. N2.

' Reaction mixture turned dark purple. indicating
dye reduction, except when boiled cells were used.

 

APPL. ENVIRON. MICROBIOL.

suspensions of B. termitidis DAD-50 fermented
UA to 002, NH;, and acetate (Table 5). Fructose
or formate was neither required for nor stimu-
latory to UA dissimilation by this strain.

Anaerobic uricolysis by Citrobacter.
Studies of uricolysis by Citrobacter strains were
hampered, because isolates exhibited limited
consumption of UA in liquid media. This was
true not only with TY-based media (e.g., see
Table 4 of reference 17), but also with richer
media such as BHIU and SUA. These observa-
tions were surprising, because Citrobacter iso-
lates readily formed prominent clear zones on
UA-containing agar plates. That UA was indeed
being consumed on such plates was verified by
assay for UA in l-mrn-diameter agar plugs, re-
moved at various distances from a single streak
line of growth. Plugs from within the clear-zone
region (i.e., O to 15 mm from the streak line;
BHIU plates; 2 days of incubation) contained no
UA, whereas plugs external to the clear zone
contained UA in amounts identical to those pres-
ent before inoculation. In addition, plugs re-
moved from clear zones and then inserted into
uninoculated plates caused no clearing in the
latter. These data argued against a diffusible,
UA-degrading enzyme being released from cells
on agar media.

The following attempts to increase UA con-
sumption in broth media were unsuccessful: (i)
incubation under an atmosphere of 90% N2/ 10‘}
H2 (vol/vol); (ii) incorporation of UA into media
as a suspension rather than as an NaOH-solu-
bilized solution; and (iii) incorporation of 0.2%
agar into broth media.

Unfortunately, cell suspensions of UA-grown
cells showed no degradation of UA at all, even
if a cosubstrate (glucose, fructose, ribose, or for-
mate), which was metabolized, was included.
This was true whether such cells were derived
from broth cultures or were scraped from the
surface of BHIU plates which they had com-
pletely cleared.

TABLE 5. Fermentation of UA by cell suspensions
of B. termitidis UAD-50"

 

nmol/100 nmol of HA fer-

 

Product men ‘ ed
CO: 338.7
NH. 391.5
Acetate 77.1
‘i C recovery 98.6
‘2 N recovery 97.9
O/R index 1.03

 

" Reaction conditions were similar to those described
in Table 2, footnote a, except that UA was the only
substrate added and the temperature was 37°C. Cells
were grown in TYFU medium before harvest.

62

VOL. 40. 1980

DISCUSSION

Major products of anaerobic uricolysis by ter-
mite gut streptococci and B. termitidis were
C02, acetate, and NH3. Limited studies with
Citrobacter isolates also revealed NH; as a ma-
jor nitrogenous product (17). Mead (12) found
that a Streptococcus Iaecalis strain of chicken
origin also formed C02, acetate, and NH; during
uricolysis, but no quantitative data were re-
ported. Whole-cell suspensions of rumen bacte-
ria degrade UA to similar products (10), but
pure culture studies have not yet been done.

Uricolysis by B. termitidis conformed closely
to the theoretical equation 1UA —. 3.5CO; +
0.75acetate + 4NH; (Table 5). By contrast,
group N streptococci showed incomplete dissi-
milation of UA unless formate (or a formicogenic
compound such as fructose) was present as a
cosubstrate in amounts equirnolar to that of UA
(Fig. 2 and 3). The role of formate appeared to
be that of a reductant: formate oxidation to CO;
by formate dehydrogenase yielded 2 H’ + 2 e‘
needed to drive uricolysis to completion. Thus,
anaerobic uricolysis by strain UAD-l (Table 3)
conformed reasonably well to the following the-
oretical equations: lformate + 1UA —o 4C0; +
lacetate + 4NH..; or 2H‘ + 2e‘ + 1UA —. 3C0:
+ lacetate + 4NH;.

We do not know the reductive step(s) in uri-
colysis by streptococci that uses formate as elec-
tron donor. However, it may be one which re-
quires low-potential electrons. since formate ox-
idation exhibits an 130' -= —0.42 V (11). If reduced
nicotinamide adenine dinucleotide oxidation [Eo'
- -0.32 V (11)] could function in this regard,
one would have expected 5 nmol of fructose
(instead of 20) to drive complete dissimilation of
20 nmol of UA (Fig. 2). Likewise, in fermenta-
tions of fructose plus UA, one would have ex-
pected less lactate production per mole of fruc-
tose than from the fermentation of fructose
alone (Table 2). However, neither was the case.

To our knowledge, this is the first report of
FDH activity in a Streptococcus. However, FDH
was observed only in cells grown on UA, and
then only in the presence of an exogenous elec-
tron acceptor (UA or an electron-accepting dye;
Table 4). The latter observation is consistent
with the apparent absence of an Hg-evolving
hydrogenase in streptococci; i.e., cells cannot use
protons as an electron sink during formate oxi-
dation. FDH of strain UAD-l showed a sensitiv-
ity to air that is characteristic of the enzyme
from other bacteria (1, 8).

Major products of DA degradation by termite
gut streptococci and B. termitidis (i.e., CO2,
acetate, and N H;) were the same as those
formed by Clostridium acidiurici and C. cylin-

ANAEROBIC URICOLYSIS BY TERMITE GUT BACTERIA

131

drosporum (6). Unlike the clostridia, however,
growth of the present isolates was not strictly
dependent on purine fermentation (17).
Whether the similarity of major products re-
flects a similar pathway for intermediary metab-
olism of UA remains to be determined. However,
it is noteworthy that xanthine, the first inter-
mediate in uricolysis by clostridia, was not used
as a substrate by growing cells of Streptococcus
UAD-l or B. termitidis DAD-5O (17), nor was it
degraded by UA-grown cell suspensions of strain
UAD-l.

Based on the results of this study, and as
discussed in the accompanying paper (17), we
conclude that uricolytic gut bacteria may well
be important to the carbon, nitrogen, and energy
nutrition of R. flavipes as well as other members
of the gut microbiota.

ACKNOWLEDGMENTS

We are grateful to Clifford Sporn for excellent technical
assistance and to Donald G. Cochran for helpful discussions.

This work was supported in part by the Michigan Agricul-
tural Experiment Station of Michigan State University and
by National Science Foundation grants BMS 75-05850 and
PCM 77-05732. C.J.P. held a National Science Foundation
predoctoral fellowship during the major portion of this study

LITERATURE CITED

1. Andreeeen, J. R., and L. G. Ljungdahl. 1974. Nicotin-
amide adenine dinucleotide phosphate-dependent for-
mate dehydrogenase from Clostridium thermoaceti-
cum: purification and properties. J. Bacteriol. 120:6—14.

2. Areta. W.. H. Kaspari. and J.-H. Klemme. 1978. Uti-
lization of purines as nitrogen source by facultative
phototrophic bacteria. FEMS Microbiol. Lett. 4:249-
253.

3. Barnes. E. M., and C. S. Impey. 1974. The occurrence
and properties of uric acid decomposing anaerobic bac-
teria in the avian caecum. J. Appl. Bacteriol. 37:393-
409.

4. Bernt. E., and H. U. Bergmeyer. 1974. D-fructose, p.
1304-1307. In H. U. Bergmeyer (ed.). Methods of en-
zymatic analysis. vol. 3. Verlag Chemie Weinheim. Ac-
ademic Press. Inc., New York.

5. Block. R. J.. E. L Durrurn, and G. Zweig. 1958. Paper
chromatography and paper electrophoresis. Academic
Press. Inc., New York.

6. Champion. A. 8., and J. C. Rabinowitz. 1977. Ferre.
doxin and forrnyl-tetrahydrofolate synthetase: compar-
ative studies with Clostridium acidiurici. Clostridiurn
cyltndrosporum. and newly isolated anaerobic uric acid-
fermenting strains. J. Bacteriol. 132:1003-1020.

. Conway. E. J. 1940. Micro-diffusion analysis and volu-
metric error. D. Van Nostrand Co., Inc. New York.

8. Enoch. H. G., and R. L. Lester. 1975. The purification
and properties of formate dehydrogenase and nitrate
reductase from Escherichia coli. J. Biol. Chem. 250:
6693-6705.

9. Hunter. A., and M. H. Givens. 1914. The metabolism of
endogenous and exogenous purines in the monkey. J.
Biol. Chem. 17:37-53.

10. Jurtabuk, F., Jr.. R. N. Doetech, and J. C. Shaw. 1958.
Anaerobic purine dissimilation by washed suspensions
of bovine rumen bacteria. J. Dairy Sci. 41:190-202.

11. Leach. P. A. 1976. Oxidation-reduction potentials. ab-

KI

132

I2.

13.

14.

15.

16.

17.

18.

63

POT RIKUS AND BREZNAK

sorbance bank and molar absorbance of compounds
used in biochemical studies. p. 122-130. In G. D. Fas-
man (ed.). Handbook of biochemistry and molecular
biology, vol. 1. CRC Press, Inc., Cleveland.

Mead. G. C. 1974. Anaerobic utilization of uric acid by
some group D streptococci. J. Gen. Microbiol. 82:421-
423.

Mortenaen. A., and A. 'flndall. 1978. Uric acid metab-
olism in the eaeca of grouse (Lagopus lagopus). J.
Physiol. 284:159-1601’.

Neish. A. C. 1952. Analytical methods for bacterial fer-
mentations. 2nd rev. National Research Council of Can-
ada. Saskatoon.

Potrikus, C. J., and J. A. Breznak. 1977. Nitrogen-
fixing Enterobacter agglomerans isolated from guts of
woodoeating termites. Appl. Environ. Microbial. 83:
392—399.

Potrikus, C. J., and J. A. Breznak. 1980. Uric acid in
wood-eating termites. Insect Biochem. 10:19-27.

Potrikus, C. J., and J. A. Breznak. 1980. Uric acid-
degrading bacteria in guts of termites [Reticulitermes
flavipes (Kollar)]. Appl. Environ. Microbiol. 40:117-
124.

Schultz. J. E., and J. A. Breznak. 1979. Cross-feeding
of lactate between Streptococcus lactis and Bacteroides
sp. isolated from termite hindguts. Appl. Environ. Mi-

19.

21.

Asst. Esvmos. MICROBIOL.

crobiol. 87:1206-1210.
Sebald, M. 1962. Etude sur les bacteries anae'robies gram-
négatives asporulees. lmprimerie Bameoud S. A., Laval.

. “emu, '1‘. D., D. C. Ellwood, and V. M. C. Ipngyear.

1979. Change from homo- to heterolactic fermentation
by Streptococcus lactic resulting from glucou lirnita-
tion in anaerobic chemostat cultures. J. Bacteriol. 138:
109-117.

Umbreit, W. W., R. H. Burris. and J. I". Staufler.
1964. Manometric techniques. Burgess Publ‘mhing Co.,
Minneapolis, Minn.

. Valentine, R. C., and R. S. Wolfe. 1960. Phosphorolysis

of carbamyl oxamic acid. Biochim. Biophys. Acta 46:
389-391.

. Vogels, G. D.. and C. Van der Drift. 1976. Degradation

of purines and pyrimidines by microorganisms. Bacteo
rial. Rev. 40:403-468.

. Wagner. R., and J. R. Andreesen. 1977. Differentiation

between Clostrt'dium acidiurici and Clostrtdtum cylino
drosporum on the basis of specific metal requirements
for formate dehydrogenase formation. Arch. Microbiol.
114:219-224.

. Whiteley, H. R., and H. C. Douglas. 1951. The fermen-

tation of purines by Micrococcus lactt'lytt'cus. J. Bacte-
rial. 61:605-616.

ARTICIEIV

GUT BACIERIA mans URIC PCID:

ammmmflmmm

C. J. Potrikus and J. A. Breznak

64

ABS'I'RPCI‘

lbticulitermes flavipes termites possess purine nucleoside
phosphorylase and xanthine dehydrogenase to synthesize uric acid (UA) ,
but lack uricase or any other means to degrade UA. 'lhis seemed
unusual, because termites do not excrete UA, but was consistent with
the low N content of termite feces. However, uricolysis does occur in
termites, but it is an anaerobic process irradiated solely by the hindgut
microbiota. Moreover, 1‘C-tracer experiments showed that termites
transport Uh firm the site of synthesis and storage (fat body tissue)
to the site of degradation (gut microbiota) via Malpighian tubules.
N33, a major product of uricolysis by the gut bacteria, is a
potential N source for termites directly through glutamine synthetase
activity present in fat body tissue. Results indicate that uricolysis
by the termites’ gut microbiota recycles significant amounts of N and
appears to be an inportant means of N conservation for these
oligcnitrotrcphic insects.

65

{v )V“

n"

9 1

, k

mmowcmm

'me ability of synbiotic microbes to augment the nutrition of
their host is well-recognized, and numerous such functions have been
ascribed to the associated gut microbiota of termites. Because
xylophagous termites are oligonitrotrophic, i.e.. their food is low in
carbined N (13) , gut microbes have frequently been inplicated in
termite N econauy. Studies by Hungate (11) indicated that termites,
with the aid of fungi, could efficiently sequester the low amounts of
combined N from wood and soil. Recent evidence also indicates that
N2 fixation by gut bacteria may contribute to N acquisition by
termites (5,24).

In addition to N acquisition, conservation of N should also be
inportant to termite N econauy and could be effected by the recycling
of nitrogenous wastes (l4) . Certain termite species form significant
ammnts of uric acid (UA) , but unlike many other terrestrial insects
they do not void the purine despite the apparent lack of uricase in
termite tissues (25). However, detailed studies of the gut microbiota
of the termite Reticulitermes flavipes revealed an abundant population
of uricolytic bacteria capable of degrading DA in pure culture to
C02: 1'33. and acetate (26,27). All of these findings are
consistent with the hypothesis of leach and Granovsky (14) who

speculated that recycling of DA by termite gut microbes was inportant

66

67

to the N nutrition of termites.

Little information is available regarding the physical and
chemical aspects of (IA synthesis, storage, mobilization, and
degradation in termites (13), or the relevance of gut microbes to these
processess _ig M. Accordingly, the present study was done to assess

the inportance of bacterial uricolysis to conservation of N in 5.

flaviEs.

MERIALSANDW

Insects and bacterial strains. mticulitermes flavipes (Roller)

 

termites were collected in Janesville, WI, Dansville, MI, and Spring
Arbor , MI and were maintained in the laboratory as previously described
(29). Worker termites (i.e., externally undifferentiated larvae beyond
the third instar) were used for all experiments.

wecimens of Drosophila melanogaster were obtained fran the
culture collection of '1‘. Friednan of Michigan State University.

Uricolytic bacteria were Streptococcus strain UAD-l and

 

Bacteroides termitidis strain DAD-50. Both strains were isolated fran

guts of g. flavig and were described previously (26) .

Preparation of tissues, cell w _a_nd cell extracts.
Termite Malpighian tubules were obtained by rennving the intestinal
tract (24) into a solution of 0.25 5 sucrose containing 1.7 all mm (93
6.9) and dissecting the tubules away fran adherent fat body tissue and
firm their attachment to the gut or rectum.

'lb obtain fat body tissue, individual termites were degutted and
the abdomen was sliced longitudinally with a scalpel. Fat body tissue
was then scraped fran the abdominal wall into 0.5 m1 of sucrose/mm
solution.

Abdomen extracts were an easily obtainable, but crude, source of

fat body tissue enzymes. Termites were decapitated and degutted, and
68

:., m,

69

the abdanens were then transferred to a snall tissue hanogenizer
containing 2.0 ml of 0.1 g Tris (hydroxymethyl) aminanethane (Tris)
buffer (pH 8.0) and approximately 10 mg of sand (Mallinckrodt). The
mixture was hauogenized for 2 to 3 min at 4°C, centrifuged at 27,000
x g for 15 min, and the supernatant fluid was used. Abdanens of 60
termites yielded approximately 16 mg proteian of extract.

Minced termite guts were prepared as previously described (26) ,
except suspensions contained five to ten gut equivalents in 500 p1 of
buffer.

Gut tissue was separated fran gut contents as follows: 20 guts
were removed under anaerobic conditions (24) and were placed in 500 pl
of 0.1 g potassium phosphate buffer (pH 7.0) . Each gut was sliced once
at the paunch with a scalpel. The suspension was transferred to a 13
um test tube and vigorously agitated for 3 min by using a Vortex type
mixer operating at mxinum speed. Gut segments were allowed to settle
for 2 to 3 min, and the supernatant fluid (containing the gut contents)
was aspirated to a separate tube. Gut sements were washed twice more
with 500 ill of buffer as just described. The supernatant fluid fran
the first wash was pooled with the gut contents, whereas that frau the
second wash was discarded. The remaining gut tissue segments were
finally resuspended in 500 ill of buffer.

Cell suspensions of Streptococcus UAD—l and _B_. termitidis UAD—SO
were prepared under anaerobic conditions as previously described (27) ,
except 0.1 g potassiun phosphate buffer (pH 7.0) was used and cells

were resuspended to a density of 108 to 109 cells/ml.

Assays _f_o_g gm activities. (1) Uricase-mediated _U_A;_
degradation. Uricase—mediated 0A degradation was assessed by measuring

70

MC-allantoin from [2-14C] UA as described by Friedman

formation of
and Johnson (6). In our system, 3. flavipes tissue preparations were
treated in an ultrasonic water bath (Mettler Electronics Corp.) for 2
min to aid in tissue disruption. Control assays employed Malpighian
tibiles fram newly emerged adults of Q. melagaster as previously
described (6) .

(ii) Nonuricase-mediated lg gradation. Nonuricase—mediated UA

degradation was assessed by measuring 1“(:02 evolution from

lac-labelled

[2—1‘C] Uh or conversion of the latter to other
coapourds. Reactions were carried out in a manner similar to that
described previously (26) , except that both aerobic and anaerobic (903
N2/10% H2) incubations were used. Samples (100 pl) of some
reaction mixtures were removed before acidification with m1; unaided
with 100 111 of a solution containing 5 In each of nonradioactive UA.
xanthine. and hypoxanthine; and 10 111 was then spotted on cellulose m
300 thin layer chrcmatography (TIC) sheets (Bridgman) . Radioactive
[z—l‘cl m (250 pmle) was also spotted as a control. m: sheets
were develcped in n—butanolmethamlzm4a-hlizo, 90:30:5z30 (by
vol). R

f
0.15; xanthine, 0.21: and hypoxanthine, 0.35. Following development.

values for the standards in this solvent system were: UA.

the TLC sheet was cut into 2 cam-wide vertical strips. and each strip
was cut into segments corresponding to the carrier sibstrate spots,
which were located by viewing the chromatogrms with short wavelength
UV light (Mineralite, Ultra-Violet Products, Inc., San Gabrial, CA) .
The remainder of each strip was cut into 1.5 cm segments. Each segment
was assayed for radioactivity as described above.

(iii) Xanthine W (_xgi. XDH was determined

I.

71

spectrophotametr ically by measuring substrate-dependent reduction of
NAD to m2 (23). Absorbance measurements were made by using a Cary
model 219 recording spectrophotometer (Varian Associates, Inc. , Palo
Alto, CA) . Reaction mixtures (1.0 m1 final vol) contained (umole) :
substrate, 0.15; KCN, 10.0: NAD(HZ), 0.5; Tris buffer (pH 8.0), 60
to 70; and approximately 1.6 mg abdomen extract protein. When inosine
was used as substrate. 5 mole KZHPO4 was also included. maction
mixtures were preincubated for 5 min at room telperature prior to
initiation by NADGIZ) addition. Data were corrected for endogenous
NAD reduction. Since K01 was included in the reaction mixtures,
m2 oxidation was negligible. For calculation of enzyme activity,
the molar extinction coefficient of m2 at 340 ram was assumed to be
6.22 x 103 (23) . One unit of enzyme activity is defined here as that
anountwhchresults inthereductionofluwleof NAD to NAM!2 x
min-1.

Reaction mixtures containing [2—14C] xanthine were also analyzed

14

for production of C-UA. This was accomplished by measuring

14(.‘.--a11antoin. After

uricase-dependent conversion of the product to
NA!) reduction had been mitored, two 0.4 ml portions were removed from
each reaction mixture; 20 ul of uricase enzyme (hog liver, Type V;
Sigma Chemical Co., St. Louis, m) was added to one portion and 20 ul
of Tris buffer to the other. After both portions were incubated at
room temperature for 2.5 h, 20 ul of each mixture was taken for TIC as

previously described (25) , except the developing solvent used was
isopropamlmlmzo (65:16.7:18.3, by vol). Allantoin was located on
chromatograms by spraying control strips of the TIC sheet with
Ehrlich’s reagent (1).

if

 

72

(iv) Glutamine synthetase (GS) . GS activity was assayed by
measuring the production of L—X-glutamylhydroxamate from hydroxylamine
and L-glutamate (synthetase activity) or L-glutamine (transferase
activity). Reactions were carried out as described by ibwe et al (28) ,
except that in some cases 0.2 ml of termite abdomen extract (equivalent
to approximately 3.2 mg protein) was used and the incubation period was
extended up to 60 min. Spectrophotometric measurements were made by
using a Varian model 6348 double beam spectrophotometer operating at
535 nm. The mount of product formed was determined by reference to a
standard curve obtained with L-X-glutamylhydroxamate in the canplete
reaction mixture. One unit of enzyme activity is defined here as that
amount which results in the production of one mole of

L-X-glutamylhydroxamate x min-1.

Uricolgis by cell suspensions g termite gut bacteria.

Uricolysis by cell suspensions of Streptococcus UAD—l and g.

 

termitidis DAD-SO were carried out as described previously for
metabolism of DA by minced termite guts (26) , except 0.1 mole of Na
formate was added to reaction mixtures containing Streptococcus cell
suspensions, and incubation was for 60 min at 25°C. Individual

reaction mixtures contained 1 x 108 to 5 x 108 cells.

We of (_1_A; g termite Malpighian tubules. The method used to
measure UA uptake by Malpighian tubules was modeled after the system
described by Sullivan and Sullivan (31) . Preparative procedures were
carried out aerobically at 4°C. Four or five guts, with Malpighian
tubules attached, were removed from _13. flavipes and placed intact into

300 pl of Grace’s Insect Tissue Culture Medium (Grand Island Biol.

73

Co., New York) in a shallow depression plate. Twenty mole of
[2-14C] DA and 0.31 mg of 3I-I--inulin (FIT! = 5200) were added to

each well, which was then covered with a glass slide to prevent
excessive evaporation. Reactions were incubated at 25°C. Inulin
(rot generally transported by Malpighian tubules because of its high
molecular weight) served as an internal control for tissue integrity.
At the end of the incubation periods, reaction mixtures were cooled to
4°C to retard transport activity. Guts were then individually
removed, rinsed briefly in 500 pl of sucrose/mm solution (above) at
4°C, and the tubules were separated from gut tissues. All sets of
tubules (8 tubules/set) from a single well were combined in 50 pl of
0.1 g acetic acid in a 1.5 cc test bibe (Eppendorf) at 4°C. The time
required for dissection of tubules from five guts was 15 min, and was
consistent for each sample, so residual transport activity at 4°C
during the preparation period should not have significantly affected
the results. Tubule preparations were extracted by sonication (see
above) for 10 min, followed by incubation at 4°C for 5 min.
Preparations were centrifuged for 5 min at 4°C in an Eppendorf mdel
5412 centrifuge (Brinkman), and the radioactivity in a 20 or 100 pl
sample of the supernatant fluid was determined.

_Ig vivo uricolysis H3; flavipes termites. .12. 31.19 uricolysis
was assessed by measuring l4C02 evolution from [2—14C] UA
injected into termite hemolymph. A solution containing [2_14C] UA
(0.5 mole/m1) and 3m-inulin (78.9 ug/ml) was injected into the
abdomen of _R. flavipes. Injection needles were made from 1.0mm o.d.
x 0.8mm i.d. pyrex capillary tubing (Drummond Scientific Co.,

Broomall, PA) drawn to a very fine point. Termites were immobilized

74

for injection by cooling them to 4°C and restraining them in a trough

(at room temperature) formed by embedding a longitudinally—sliced piece
of Teflon tubing (1 mm o.d.) in a sheet of soft clay. Clay was pressed
gently against the olbing to fold the termite in position. Injections
were made while viewing the termites under a dissecting microscope.
The needle was inserted into the ventral side of the abdomen, just
beneath the cuticle, between the fifth and sixth sternites.

3

Preliminary tests with H 0 standards indicated that approximately

2
0.2 pl of solution could be injected into individual termites without
apparent injury to the insects. This volume of solution caused an
apparent swelling of the abdomen, but did not result in leakage of
fluid from the injection site. The mount of solution injected was
thus gauged by observing the extent of abdominal swelling.

Following injection, termites were placed in 5 ml capacity serum
vials which were then sealed with rubber stoppers. Usually five
termites were placed in a single vial. Vials were placed at 25°C,
and at regular intervals each one was flushed with air for 10 min, at a
rate of about 20 cc/min. The exit air was bubbled directly into a
solution of Aqueous Counting Scintillant (Mersham Corp., Arlington
Heights, IL) :phenethylamine:methanol (10:3:3, by vol) to trap

14C102. The trapping efficiency of this procedure was determined to

be 97.9%. Measurelent of 14

C-radioactivity was made as described
below.

After incubation, termites were removed from vials and cooled to
4°C. alts were reioved, extracted with 0.1 N acetic acid (see
above), and the 3H--radioactivity present in extracts was determined.

Samples in which 3H—-radioactivity appeared in the gut extracts

75

indicated that the gut may have been pierced during injection and thus

were rot included in results.

14

To determine whether CO was evolved from excreta deposited

2
in the vials during the incubation period, vials were resealed after
retoval of the termites and were incubated at 25°C for one additional
hour. After incubation, vials were flushed and 14(1)2 was trapped

as described above.

Other procedures. UA content of termites was determined
enzymatically after extraction of [IA from tissues with 1.12003 (25) .
Protein content of termite extracts was determined by the biuret method
(9) with bovine serum albumin as standard. Radioactivity measurerents
and quench corrections were made as previously described.
14C-toluene and 3820 were used as standards. Enumeration of
uricolytic bacteria in guts of _13. flaviEs was done by quantitative

plating on SUA medium as described previously (26) .

3

Chemicals. [2—14C] HA and H-inulin were obtained from

Amershom 00:13.; [2-14C] xanthine from msearch Products

14C-(rioluene and 3H20 from New

International, Elk Grove, IL: and
England Nuclear, Boston, MA. Scintillation grade phenethylamine was
obtained from Eastman Kodak Co., mchester, NY. mrines, Li2003,
Tris, NAD(H2) , ATP, ADP, L—glutamine, L-glutamate,
L-X-glutamylhydroxamate (L-glutamic acid X-monohydrate) , and bovine
serum albumin were from Sigma Chemical Co. All other commercially

available chemicals were of reagent grade.

REUL'I‘S

p_r_ig acid mthesis by termite tissue extracts. 3. flaviEs
tissues possessed key terminal enzymes of (IA synthesis. Abdomen
extracts readily reduced NAD to NAIXi2 in the presence of xanthine or
hypoxanthine, suggesting the presence of XDH. The reaction was
dependent upon the presence of substrate and extract in the reaction
mixture. No NAD reduction was observed when boiled (5 min) extract was
used. XDH activities calculated from the initial linear portion of the

4 and9.30x

reaction were (units x mg protein-1) : 3.69 x 10-
10"4 for xanthine and hypoxanthine, respectively.

Purine nucleoside phosphorylase also appeared to be present in
abdomen extracts. Inosine served as a substrate for NAD reduction, but
only in the presence of added phosphate. However, rates of
phosphorylase activity were difficult to determine precisely, due to
relatively high endogenous rates of NAD reduction with no added
inosine. Since inosine is present in termite extracts (25) , it
presumably served as substrate when exogenous phosphate was added.

To further confirm the UA-synthesizing ability of termites, and to
assure that NAD reduction with xanthine by abdolen extracts represented
UA formation, an examination was made for DA as a product of XDH
activity. The lack of uricase in termite tissues (see below) and the

use of 14

14

C-xanthine as substrate aided this examination. Any

14

C-UA formed was converted to C-allantoin by reaction with

76

77

commercial uricase, and the 14C-allantoin was separated from

1“C—xanthine by TTC for determination of radioactivity. With the TLC

system used, xanthine and DA spots overlapped so their combined
radioactivity was tabulated. Ibvever, the critical product (allantoin)
was well separated from the substrates (e.g., Rf for allantoin =

0.53; Rf

presented in Table l. Radioactive UA (detected after conversion to
14

for HA 8 0.38) and could be easily assayed. Results are

C-allantoin with commercial uricase) was readily formed from

14C--xanthine by abdomen extracts in the complete reaction mixture.

l4C-allantoin was formed if this reaction mixture was

14

As expected, no

not treated with commercial uricase. Little C—UA was formed if the

reaction mixture lacked added NAD or contained boiled extract (Table
l) . Lack of HA formation when NAD was omitted strongly suggested that
oxidation of xanthine to 0A was mediated by XDH rather than xanthine

oxidase .

A control reaction mixture containing l4C-UA instead of

14C-xanthine verified that commercial uricase was active under the

14

assay conditions used, since 96% of the C-UA was converted to

1l‘C-allantoin (Table 1) . This conversion was also accompanied by a
marked decrease in the absorbance of the reaction mixture at 292nm (the
absorption maximum of HA; data not sham) . As expected, reaction

14c-allantoin,

mixtures not treated with uricase yielded no
supporting our previous conclusion that termite tissues lack uricase

(see below) .

Lack 9g uricolytic activity in 3°. flaviEs tissues. 5. flavipes
termites do rot possess uricase. This was first suggested by

spectrophotometr ic assays with crude tissue homogenates (25) and is now

8
7

saw: Ho

IN. mmHoea\Hun we .oofifiucmx _o
MHuxm coEoopm oom mm<z “ucmmco m

.COfiumcHEhouoo o3u mo some ecu who can .muocfia Moo moowumuwmuowmeo o

INH "moeuw>«ucm oemwooom

.cae m “cocoao

.oaoeaeeu: mm .oaon

.moosumz can mamfiuoumz ow ooumuwoce
wwe “20M “moasucmx HUQHINH ”oocfimucoo oucuxHE coaucmou ouoHoEoo

a

on omumaoomu one ounce

 

 

 

 

 

«H
Hmnmma mwomma Nmmqoa nmqmoa mmchH mmmmma NNmQHH nomoaa >uw>aucn0fiomm
Hence
oawnma o mNm qu moo o Nonm o oeoucmaa<
mcwsucmx
Hmmm mwomma nooooa nqmmoa wmmmaa mmmmma ooamoa nomcaa +
ween cap:
enough: canoes: convex: convex: mmmcfiu: canoes: wooden: omnofiuc woman way
cues one“: cow: was“: :uH3 maze: Suez once: so mooN
owes canolu moan uumuuxo omHHon
mmcfixucmxioewamoowz u “muoHoEoU m<z mace: muoHoEoo
nmuouxwz :owuomom
mmuunuuxm beacons mome>nam .m.%o oofisuomxloqa Eouw oowuoooouo ofiom 3.310.»H .H canny

79

confirmed by more sensitive radiometric assays (Table 2). Neither

Malpighian tubule tissue tor fat body tissue converted [2-14C] UA to

l4C—allantoin. This was true whether termites used were low in DA

stores (i.e., 6.3% UA, dry weight basis) or contained over 20% UA.
Control Malpighian tubules from Q. melanogaster readily converted

14C-UA to 14C-allantoin (Table 2) . Moreover, when tubules from g.

flaviEs and _D_. melanogaster were mixed together, the mount of
allantoin formed was oorparable to that formed by Q. melanogaster
tubules alone (Table 2) . This result indicated that termite tissues
(i) did rot inhibit uricase activity, and (ii) did not contain an
allantoin-degrading enzyme system.
Termite tissue extracts were examined for their ability to degrade
(2—14C] UA to compounds other than allantoin. Abdoten extracts
14

failed to evolve (1)2

aerobically. Moreover, TTC of reaction mixtures incubated with

l4C-rad ioactivity

from [z-l‘cl um when incubated

[2-1‘c1 GA for 3 h showed no redistribution of
with colpared to reaction mixtures lacking termite abdolen extracts
(average 14C recovery from TIC plates was 96.13) . From the data of

Tables 1 and 2 we concluded that _R. flaviEs tissues can synthesize
UA, buttheycannotdegrade UA nor can they convert UA to other

compounds in significant mounts.

Bacterial uricolysis _ig 9% of R_. flaviEs. If termites are to
recover N from UA _i_rl gilo, some ronuricase—mediated degradation of DA
must occur. Initial studies with 3. flavigs revealed that
significant amounts of 14(202 were evolved when minced termite guts
were incubated with [2—14C] UA anaerobically (26) . We therefore

sought to confirm and extend these observations. Minced guts prepared

80

Table 2. Absence of uricase in tissues of _R. flavipesa

 

 

 

 

 

UA content of 14C-Allantoin
Preparationb insects (_% d_ry_ _wt_._)_ formed (gm)
3. flaviEs M.T. 6.3 0
F.B. 6.3 0
_B_. flaviEs M.T. 23.6 0
F.B. 23.6 0
_Q. melanogaster M.T. N.D.C 19,125
E. melanogaster M.T. N.D.
+ 18,573
_R. flaviEs M.T. 6.3

 

aReaction mixtures (25 ill) contained 4 mole [2-14C] UA (sp. act.

= 57 uCi/lmmole) . Incubation: 5 min at 26 C.

bM.T. = Malpighian tubules; F.B. = fat body tissue. One set of 8
M.T. from g. flavipes or a single tubule from _Q. melanogaster, or
F.B. from one _R. flaviEs termite was used for individual reactions.

 

cN.D. = not determined.

81

from laboratory-maintained termites evolved 14(:02 from [2-14C] UA

14(1)2 evolved x gut".1 x h"1 (Spring

Arbor termites) and 82.4 pmole 1“(1)2 evolved x gut-1 1

at rates of 27.7 pmole
x h—
(Dansville termites), but only under anaerobic incubation (Table 3).
Moreover, virtually all of the uricolytic activity of termite guts was
associated with the gut contents: little or no ur icolysis was
catalyzed by gut tissue alone (Table 3). As expected, uricolysis by
gut contents was an anaerobic process (Table 3) .

Minced gut preparations from field specimens of _R. flaviEs also
exhibited uricolysis anaerobically, and rates of uricolysis were
comparable to those of laboratory-maintained termites (Table 3) .

If uricolytic activity of termite gut contents was to be
attributed to gut bacteria, pure cultures of such bacteria (26,27)

14

should also be capable of evolving C02

proved to be the case. Streptococcus UAD—l and _B_. termitidis UAD—SO

readily evolved 14(1)2 from [2-14C] UA under anaerobic conditions.

from [2-“c1 UA. This

 

Cell suspensions of Streptococcus UAD-l grown on fructose plus UA

evolved l4(1)2 at rates of up to 249 pmole x (108

h-l, whereas no “(132 was evolved by cells grown on fructose

alone. 2. termitidis DAD-50 cells grown on fructose evolved 14(1)2

from [2—14C] UA at a rate of 2049 pmole x (108 cells)’1 x h'l,

cells) '1 x

 

but only after a 30 min lag period. Presumably, the uricolytic enzyme

system of strain DAD-50 was induced during this period. Cells of

strain DAD-50 grown on fructose plus UA evolved l4C02 at a faster

rate (6734 plole x (108 cells)-1 x h-l), with no apparent lag
period. No 1«(1)2 was evolved by either strain when incubated

aerobically.

82

14 . 14 . . .
Table 3. CO2 evolution from [2- C] uric ac1d by 3. flalees

gut preparations

 

 

 

14C02 evolutione
Origin of Collection Gas pmoles
a b . c d
specimens site Preparation phase gut equivalent x h
Laboratory D Minced guts Air < l (2)
N2/H2 82.4:l.8 (2)
Cut contentsf Air < 1 (5)
Nz/H2 57.1:3.8 (21)
Cut tissue N2/H2 < l (2)
SA Minced guts Air < l (l)
T
hz/HZ 27.7 (1)
Field D Minced guts NZ/HZ 36.1:3.4 (5)
J Minced guts Nz/H2 29.6i2.0 (3)

 

aLaboratory specimens had been maintained in the laboratory for 9 or 10
months prior to assay. Field specimens were assayed within 24 h of
collection.

bD = Dansville, MI; SA = Spring Arbor, MI; J = Janesville, WI.

cPreparations contained 5 or 10 gut equivalents (minced guts or gut
contents) or 20 gut tissue equivalents. Incubation was at 25°C and
80 osc/min for 0.5 to 3.0 hours.

dN2/H2: 90/10, vol/vol.

Values are given as mean i_standard error where possible. Number
parentheses indicates the number of experiments. Sp. act. of [2-
UA = 50 pCi/umole.

11
13C]

fPopulations of uricolytic bacteria were approximately 1 x 105 viable
cells per gut.

83

These data indicated that UA degradation can occur in termites,
but it is an anaerobic process carried out by the gut bacteria and not

by the termite tissues.

Tranggrt ft_J_A; from termite fat body _t2 the gut. It was

 

 

important to determine whether termites could transport BA from the
site of its synthesis (i.e., the fat body tissue, which lies external
to the alimentary tract) to the site of its degradation (the gut
microbiota) . In most insects, transport of urinary compounds to the
gut occurs via the Malpighian tubules. Accordingly, two experimental
approaches were used: an _i_n_ !_i_t£_o_ one which measured [2-14C] UA
uptake by termite Malpighian tubules; and an in 2.1.19 one which
measured degradation of [2-14C] UA initially injected into termite
hetolymph.

The ability of termite Malpighian tubules to take up [2-14C] UA
is depicted in Figure 1. The initial rate of uptake was equivalent to
about 0.1 mole 14mm x (tubule set)"1 x min'l. After 90 min,
radioactivity equivalent to about 3 pmole (2-140] UA was associated
with each set of tubules. Continued incubation up to 10 h did rot lead
to further accumulation of radioactivity (data not atom) . Maintenance
of tissue integrity was verified by the exclusion of 3H--inulin from
those tubules which were actively taking up UA (Figure l). Boiled
tubules accumulated no 14C-radioactivity.

When [2-14C] was injected into the helolymph of live termites,
MC132 was readily evolved (Figure 2) . Clearly, the UA must have
been transported to the gut contents since that is the sole site of

14

uricolysis in termites. The initial rate of (:02 evolution, which

was an estimate of the rate of HA degradation by the gut microbiota _i_n

84

Figure l. Uptake of [2-14C] uric acid by Malpighian tubules of g.
flaviEs. Data points represent radioactivity taken up by
eacbqset of eight individual tubul 3. Specific activities:
[2— C] uric acid, 50 uCi/umole; H—inulin, 577 pCi/mg.

85

our

mots—:5.

o co

 

 

 

 

n ........ .. is). ..... +4-...

53:. - In
Aozv oozom

 

 

    
 

 

 

O
O
N
)as emqnuwdp

cow

Figure 2.

86

FNolution of 14002 from [2-14C] uric acid injected

into the helolymph of 3. flaviEs. Each point represents
the mean of four sample vials, each containing five
termites. Bars repres at the standard error of the mean.
Specific activity of [2— C] uric acid was 53 nCi/mole.
Termites were laboratory specimens collected from Dansville,
MI ten months prior to assay.

87

 

 

 

 

 

300m. selowd

Figure 2.

88

situ, was about 1 prole x termite-l x h-l. In separate

experiments, initial rates of UA degradation ranged up to 11.7 pmole x

termite.1 x h-l. Importantly, test vials from which injected
termites had been subsequently removed failed to yield l«€02 with
14

further incubation. This argued against the possibility that CO2
was being evolved from termite excreta.

The mount of 14

CD2 evolved in an assay vial was proportional
to the number of termites held therein. For example, after 2 h of
incubation, vials containing one, two, three, and four termites
contained 12.5, 35.2, 46.6, and 72.1 {moles “002, respectively.
These data verified that UA transport and degradation occurs within
each termite in the population.

5. flaviEs termites freshly collected from the field

(Janesville, WI) also evolved 14(I)

14

2 when injected with [2—14C]

UA. Rates of CD2 evolution were- comparable to those observed
with laboratory-maintained termites, and ranged from 5.8 to 9.1 pmoles
x termite"l x h-1 for the first tour of incubation.

Glutamine synthetase activity _i_n_ abdomen extracts. Glutamine
synthetase activity was present in termite abdomen extracts.
Activities were (units x mg protein'l): synthetase, 7.16 i 1.46 x

10", transferase, 8.45 i 0.60 x 10-3. Each value represents

the mean t the standard error of the mean for four determinations
made using samples of a single termite extract preparation. No
activity was observed in the absence of added houogenate, ATP, or
L-glutamate (synthetase assay), or in the absence of halogenate,
L-glutamine, Mind”, or phosphate (transferase assay). Crude extracts

which were stored at -20°C for two weeks showed a lO-fold reduction

89

in transferase activity. Levenbrook and Kuhn (16) reported a similar
instability of glutamine synthetase prepared from the Southern armyworm

(Prodenia er idania) .

 

DISCUSSIQJ

This paper reveals a new dimension to the interactions of termites
with their intestinal microbiota: that of microbial-mediated recycling
of a termite nitrogenous excretory product. 3. flaviEs termites
synthesize and store UA, but they cannot degrade the purine by
themselves. Instead, they transport UA to the alimentary tract where
anaerobic uricolysis occurs via the hindgut microbiota. This allows
recycling of the atoms of UA back to the termite. A particularly
important consequence of the symbiosis is the conservation of N within
the termite colony.

The presence of XDH in termite abdominal extracts was not
surprising, since termites can accumulate high levels of UA in fat body
tissue (25) . xvii-catalyzed reductions of hypoxanthine to xanthine and
xanthine to DA are terminal steps in UA synthesis by both the
nucleicolytic and uricotelic pathways (3) . The fact that the termite
enzyme appears to be a dehydrogenase rather than an oxidase correlates
with findings made with other insects (3,8,12,23). Moreover, terminal
oxidation of NADH2 formed by XDH activity may make UA synthesis by
termites a metabolically inexpensive means by which to store nitrogen
(3) .

Levels of observed XDI-I activity in crude extracts of _13. flaviEs
were low cotpared to those observed in Q. melanogaster (10,23) and

 

some other insects (8,12). Nevertheless, these levels can fully
90

91

account for observed rates of UA accumulation by g. flavipes. For

4

example, XIII activity was 3.69 x 10' units x mg protein-1 when

xanthine was used as substrate. Since the average protein content of

termites used for the assays was 0.4 mg x termite-1, this translates

.4 units per termite (i.e., 1.48 x 10'”4 moles UA

synthesized x termite"1 x min-1) or 6.5 nmoles UA x termite"1 x

month'l. The highest rate of accumulation of HA we have observed for

1

to 1.48 x 10

_13. flaviEs was 0.027 mg UA (8 0.16 mole) x termite"1 x month-
(25) , i.e., a rate only 2.5% that of observed XDH activity. Therefore,
Km activity would appear tot to limit UA synthesis in _R_. flaviEs.

The apparent presence of purine nucleoside phosphorylase activity
in abdomen extracts was consistent with our previous finding of inosine
in extracts of _R. flaviEs (25) and with studies with other insects
(3,4,10) .

Novel (i.e., nonuricase-mediated) pathways for uricolysis have
been suggested for insects such as Per iplaneta amer icana (18) and
Anthonoms (21) . However, whereas 3. flaviEs could synthesize and
store UA, they could not alone degrade it. This was concluded from the
absence of uricase in termite tissues and from the inability of termite
tissues or extracts to effect: (i) a decrease in the 292 mm absorption
maximum of UA in solution; (ii) a change in the TIC profile of
[2—14C] UA as colpared to an untreated [2-14C] UA control; and
(iii) 14(:02 evolution from [2-14C] UA.

14

(1)2 evolution

from [2-14C] UA, this activity was used to evaluate uricolysis by the

Since termites themselves could rot mediate

gut bacteria in vitro and _i_n vivo. The legitimacy of this approach was
validated by the ability of uricolytic bacterial isolates (26,27) to

92

evolve 14632 from [2—14C] UA. Uricolysis in _R. flaviEs was
confined to the gut contents and was strictly an anaerobic phenomenon
(Table 3) . The latter finding was consistent with on previous studies
of uricolysis by these bacteria (26,27) .

Rates of anaerobic uricolysis by termite gut preparations varied
somewhat depending upon the termites used (Table 3) , but were
significantly higher than rates previously reported (26) . Since we do
not yet know what factors (e.g., nutritional, environmental) influence
the population of uricolytic bacteria in termite hindguts (see 26) or
the rates of uricolysis by gut preparations, it is difficult to assess
the significance of these differences. Important to note, however, is
that gut preparations from both laboratory-maintained termites and
freshly collected field samples exhibited high rates of anaerobic
uricolysis. Since uricolysis by pure cultures of gut bacteria was an
inducible activity (see Results, and 27) , the ready evolution of
14(1)2 from 14C-UA by gut preparations suggests that UA transport
to the termite gut is a continuous process.

Since termites do not void UA (25) , bacteria presumably degrade
any UA elaborated into the gut (see below). Indeed, rates of
uricolysis (estimated by manometry) by pure cultures of termite gut '
bacteria correlated favorably with rates of uricolysis by gut
preparations (26,27) . Similar colparisons may be made by using rates
of 14002 evolution from [2-14C] UA by gut bacteria and from the
data of Table 3. For example, UA-grown cells of _B_. termitidis HAD-50

14(1)2 at a rate of 6734 pmoles x (108 cells) '1 x h-l,

 

evolved
suggesting that about 8 x 105 l_3_. termitidis cells per gut would be

necessary to accolnt for the activity observed with gut contents (Table

93

3) . The predicted population level colpares well with that determined

by plate counts (1 x 105 cells/gut; Table 3, footnote f). The rate

14
of CD2

pmoles x (108 cells)-

evolution from [2-14C] UA by Streptococcus UAD-l (249
l

 

x h-l) was rather low cotpared to rates
previously observed (27) : perhaps cells used in the present experiment
were not fully induced.

Uricolysis by termite gut preparations _i_n_ gitfl would be of
questionable significance _i_n g_i_v_9_ unless UA from the fat body or
hemolymph could reach the hindgut for microbial degradation. Termite
Malpighian tubules can mediate such transport (Figures 1 and 2). _R_.
flaviEs possesses eight Malpighian tubules which insert at the
proctodeal valve, just anterior to the hindgut (22). Althoigh no
detailed studies have been made on the physiology of excretion by
termite Malpighian tubules, UA transport wolld presumably occur in a
manner similar to that described for numerous other insects (e.g., see
3,17), i.e., it would pass into the tubules from the heiolym'ph and be
carried to the hindgut by the natural flow of urine through the tubules
(l7) .

The rate of DA transport by termite tubules in 3332 [about 400
dpm (= 3.7 moles) x (set of tubules).l x h.1 (Figure 1)] would
appear to be more limiting to gut uricolysis than rates of microbial
uricolysis (see above). In fact, i_n y_i_t_r_g rates of DA uptake by
Malpighian tubules (Figure l) correlated well with rates of in 1119.

uricolysis by termites injected with 14

C-UA (Figure 2). Moreover,
the fact that freshly-collected field specimens of g. flaviEs
exhibited significant rates of uricolysis suggests that UA synthesis,

transport, and catabolism are comon processes in the termite’s natural

94

environment.

The hypothesis presented by leach and Granovsky (l4) eiphasized N
conservation in termite colonies by assimilation of UA-N into microbial
protoplasm, which would then remain in the colony as a result of the
insects’ social habits (l4) . ibwever, the presence of glutamine
synthetase activity in termite tissues suggests that N33 produced by
gut microbial uricolysis (27) may be used directly by the termite.
Glutamine synthetase mediates synthesis of glutamine from glutamate and

NH (15,19) . Glutamine formed as a product of the reaction may

3
subsequently be used for numerois biosynthetic reactions (15,19) .
Glutamate is found in _R. flaviEs (2). Whether the termite competes
with its gut microbiota for any free N113 produced by uricolysis, and

then uses that NH for its own N needs, is not yet known. However,

3
even if UA-N is used directly only by the gut microbes, it coild still
be conserved within the colony, as mentioned above, by the
trophallactic and coprophagous habits of termites (13) .

The intricacies of N flow in this regard remain to be defined, but
the contribution of bacterial uricolysis to the N economy of termites
is undoubtedly significant. For example, the uricolytic activity of
guts reloved from freshly-collected termites (Dansville) was about 36

l l

x h- (Table 3). If an average colony of _R_.

flaviEs contains 6 x 104 worker termites (7) , then in the period of

pmoles x gut-

one year the gut microbiota could release more than 1.0 g of UA-N.
Since the average N content of 5. flavipes workers is abolt 0.1 mg N x
termite-1 (25) , this amount of N is equivalent to more than 10,000

termites, or nearly 20% of the total worker population. Obviously,

many variables (nutritional, environmental, tormonal) could effect

95

these rates. However, the conclusion seems inescapable that gut

microbial uricolysis is relevant to N conservation in termites.

WIS

We are grateful to D. G. Cochran for advice and encouragement,
and to T. Freioman for help with uricase assays and advice on
microinjection techniques. This research was supported in part by the
Michigan Agriculture Experiment Station of Michigan State University
and by NSF grants P04 77-05732 and P04 80-12026. C. J. P. held an

NSF predoctoral fellowship during a portion of this study.

96

10.

11.

LITERATURE CITED

Block, R.J., E.L. Durrum, and G. Zweig. 1958. A manual of paper
chromatography and paper electrophoresis. Academic Press, New
York.

Carter, F.L., and R.V. Smythe. 1973. Effect of sound and
Linzentes - degraded wood on the amino acid corposition of
Reticulitermes flavipes. J. Insect Physiol. §:1623-l629.

Cochran, D.G. 1975. accretion in insects, p. 178-281. In D.J.
Candy and B.A. Kilby (eds.), Insect biochemistry and
function. Chapman and Hall, Icndon.

Cochran, D.G., and C.F. Bruno. 1963. A partial pathway for the
synthesis of uric acid in the American cockroach. Proc. XVI
Internat. Congr. 2001., no. 2, p. 92.

French, J.R.J., G.L. Turner, and J.F. Bradbury. 1976. Nitrogen
fixation by bacteria from the hindgut of termites. J. Gen.
Microbiol. _9_5:202-206.

Friedman, T.B., and D.H. Johnson. 1977. Telporal control of
urate oxidase activity in Drgflhila: evidence of an
autonoIous timer in Malpighian tu les. Science 121:477—479.

Haverty, 14.1. 1976. Termites. Pest control 22:12-17, 46, 47, 49.

Hayashi, Y. 1961. On the role of diphosphopyridine nucleotide
(DPN) in xanthine-oxidation by the xanthine dehydrogenase of
silkworm,§o_day_x mori. Jap. J. Appl. B‘ntolol. Zool.
5:282-283.

Herbert, D., P.J. Phippsr and R.E. Strange. 1971. Chemical
analysis of microbial cells, p. 209-344. In R. Norris and
D.W. Ribbons (eds.) , Methods in microbiology. vol. 58.
Academic Press, New York.

30699: L.D., and E. Glassman. 1967. Purine catabolism in
Drosophila melanogaster. I. Reactions leading to xanthine
dehydrogenase. Biochim. Biophys. Acta L42 : 335-343.

Hungate, R.E. 1941. Experiments on the nitrogen economy of
termites. Ann. EntoIol. Soc. Am. 34:467-489.

97

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

98

Irzykiewitz, H. 1955. Xanthine oxidase of the clothes moth,
Tineola bisselliella, and sole other insects. Aust. J.
biol. Sci. _8_:369-377.

LaFage, J.P., and W.L. Nutting. 1978. Nutrient dynamics of
termites, p. 165-232. In M.V. Brian (ed.), Production
ecology of ants and termites. Cambridge Univ. Press,
Cambridge.

leach, J.G., and A.A. Granovsky. 1938. Nitrogen in the nutrition
of termites. Science 81:66-67.

Iehninger, A.L. 1975. Biochemistry. Worth Publishers, Inc., New
York.

Ievenbrook, L., and J. Kuhn. 1962. Properties and distribution
of glutamine synthetase in the Southern armyworm, Prodenia
eridania. Biochim. Biophys. Acta 63:219-232.

Maddrell, S.H.P. 1971. The mechanisms of insect excretory
systems, p. 199-331. _IgJ.W.L. Beament, J.E. Treherne, and
V.B. Wigglesworth (eds.), Advances in insect physiology, vol.
8. Academic Press, New York.

McEnroe, W. 1966. In-vivo preferential oxidation of ureide carbon
no. 2 of uric acid by Periplaneta americana. Ann. Eht. Soc.
Amer. 5_9_:1011.

Meister, A. 1962. Glutamine synthetase, p. 443-468. In P.D.
Boyer, H. Lardy, and K. Myrback (eds.), The enzymes, vol.
6. Academic Press, New York.

Meister, A., A.L. Ievintow, R.E. Greenfield, and P.A.
Abendschein. 1955. Hydrolysis and transfer reactions
catalyzed by w-amidase preparations. J. Biol. Chem.
_21_5_:441-460.

Mitlin, N. and G. Wiygul. 1973. Uric acid in nucleic and amino
acid synthesis in the boll weevil, Anthonoms grandis. J.
Insect Physiol. 19:1569-1574.

Noirot, C., and C. Noirot-Timothée. 1969. The digestive system,
p. 49-88. E K. Krishna and F.M. Weesner (eds.), Biology
of termites, vol. 1. Academic Press, New York.

Parzen, S.D., and A.S. Fox. 1964. Purification of xanthine
dehydrogenase from Droscflila melanogaster. Biochim.
Biophys. Acta 22:465-471.

Potrikus, C.J., and J.A. Breznak. 1977. Nitrogen-fixing
Enterobacter lonerans isolated from guts of wood-eating
termites. Appl. EnVlron. Microbiol. 33:392-399.

Potrikus, C.J., and J.A. Breznak. 1980. Uric acid in wood-eating
termites. Insect Biochm. .]._0_ :19-27.

26.

27.

28.

29.

30.

31.

99

Potrikus, C.J., and J.A. Breznak. 1980. Uric acid-degrading
bacteria in guts of termites [Reticulitermes flavipes
(Kollar)]. Appl. Environ. Microbiol. 10:117-124.

 

Potrikus, C.J., and J.A. Breznak. 1980. Anaerobic degradation of
uric acid by gut bacteria of termites. Appl. Environ.
Microbiol. fl :125-132.

lbwe, W.B., R.A. lbnzio, V.P. Wellner, and A. Meister. 1970.
Glutamine synthetase (sheep brain), p. 900-910. in H. Tabor
and C.W. Tabor (eds.), Methods in enzymologyv vol. 17.
Academic Press, New York.

Schultz, J .E., and J .A. Breznak. 1978. Heterotrophic bacteria
present in hindguts of wood-eating termites [Reticulitermes
flaviEs (Kollar)]. Appl. Environ. Microbiol. 33:930-936.

Schultz, J.E., and J.A. Breznak. 1979. Cross-feeding of lactate
between Streptococcus lactis and Bacteroides sp. isolated
from termite hindguts. Appl. Environ. Microbiol.
2:1206-1210.

 

Sullivan, D.T., and M.C. Sullivan. 1975. Transport defects as
the physiological basis for eye color mutants in Drosophila
melanogaster. Biochem. Genet. 13:603-613.

 

APPE‘IDIX I

WWW

APPENDIXI

WWW

Based on the results of the present study, a working model is
proposed in Figure 1 for the roles of UA and uricolytic gut bacteria in
termite N nutrition. Dietary N is incorporated into termite nucleic
acids and proteins or into gut microbe cell material. UA, usually
considered a nitrogenous waste product of terrestrial insects, is
synthesized in termite tissues from the products of nucleic acid and
protein catabolism. The terminal steps of UA synthesis are catalyzed
by nucleoside phosphorylase (NP) and xanthine dehydrogenase (XDH) .
Althoigh synthesized and stored in fat body tissue, DA is presumably in
a dynamic state within the insect, and therefore also present in the
helolymph. Termite Malpighian tubules transport UA from the helolymph
to the gut, where it is subsequently fermented by the symbiotic hindgut
bacteria to (I) , acetate, and NH3. Acetate derived from UA enters
the large pool of acetate derived from microbial cellulose fermentation
and thus serves as a carbon and energy source for the termite. NH3
is assimilated by the gut microbiota and, throigh the activity of the

termite’s glutamine synthetase, by the termite itself. Microbial N may

100

Figure 1.

101

A strategy for microbial-mediated conservation of coloined
nitrogen in termites. Solid arrows represent reactions or
processes which are known to occur as depicted. Broken
arrows depict pathways which are presuled to occur in
termites, although the specific enzymes involved have rot
been investigated. Reactions mediated by termite enzymes
are shown in the shaded portion of the figure, whereas
reactions mediated by gut microorganisms are showm in the
unshaded portion. NP, nucleoside phosphorylase; xoi,
xanthine dehydrogenase.

     

.23. A smrrcv’ron MICROBIAl é MEDIATED CONSERVATION OF
cowameo NITROGEN IN TERMITES

DIETARY

   

 

 

:Z5NITROGEN Trophallaxis
I.
‘ Coprophagy
vvvvvvvvv ‘3. H ‘ Microbial
Nucleic Acids ‘-.-t.~ g, ,. g, CellMaiarial
...... ‘- . , .
+
Acetate
+
................ 002
Malpighian Tubulc
- U . . % UricAcid
(Fat body; homolyrnph) g (Hindgui)

(Aerobic) Gui Microorganisms
(Anaerobic)

Figure 1.
102

103

also be conserved by termites through trophallaxis and coprophagy, as
well as by the ultimate utilization of nitrogenous compounds released
within the gut upon death and lysis of the symbionts.

The mechanisms for N conservation depicted in Figure l are
energetically favorable to the termite. If inosine is synthesized in
termites by the same pathway which has been documented for other
organisms, then eight ATP would be required for the production of one
inosine molecule. However, the oxidation steps catalyzed by XDH would
generate two molecules of NADH2 for each molecule of inosine
converted to UA (and thus a potential of six ATP), reducing the cost of
UA synthesis to two ATP for each UA molecule. Moreover, acetate
produced in the gut by microbial uricolysis could be oxidized by the
termite to provide additional ATP, negating the cost of UA synthesis to
the insect. In addition, uricolytic gut bacteria generate energy for
their own benefit by UA fermentation. The uricolytic bacteria
investigated in this study were not dependent upon UA for a source of
C, N, or energy. Alternative metabolic processes carried oit by these
bacteria may contribute to other aspects of termite hindgut ecology.
However, the ability of the uricolytic bacteria to acquire energy by UA
fermentation could help these microbes retain their niches in the
hindgut, as well as serving an important function in termite N
nutrition. Although the efficiency of the termite’s utilization of
acetate and NH3 produced by gut microbial uricolysis, and the expense
of Malpighian tubule transport of HA and termite NH3 assimilation are
not known, both the termite and its gut microbiota would appear to
benefit from the proposed scheme for UA metabolism. Clearly, the

synthesis and subsequent microbial-mediated catabolism of UA could be

104

an advantageous strategy for termite N conservation.

APPENDIX II

NITW FIXIM W W
ISOLATED mm GJTS OF

KID-EATIM; TEENITES

C. J. Potrikus and J. A. Breznak

Reprinted from Appl. Environ. Microbiol. 3_3: 392-399 (1977)

Anuso AND varaowulw‘nu. Micnosiowcr. Feb. 1977. p. 392-399

Copyright C 1977 American Society for Microbiology

105

Vol. 33. No. 2
Prmlcd in U.S.A.

Nitrogen-Fixing Enterobacter agglomerans Isolated from

Guts of Wood-Eating Termites‘
c. J. pormxus .wa JOHN A. BREZNAK‘

Department of Microbiology and Public Health, Michigan State University. East Lansing, Michigan 48824

Received for publication 12 August 1976

Two strains of facultatively anaerobic, N.,-fixing bacteria were isolated from
guts of Coptotermes formosanus and identified as Enterobacter agglomerans.
The deoxyribonucleic acid base composition of isolates was 52.6 and 53.1 mol%
guanine plus cytosine. Both isolates and a known strain of E. agglomerans
carried out a mixed acid type Of glucose fermentation. N2 fixation by E. agglom-
erans was inhibited by 02; consequently, N2 served as an N source only for cells
growing anaerobically in media lacking a major source of combined N. However,
peptone, NH.Cl. or KNO, served as an N source under either aerobic or
anaerobic conditions. It was estimated that 2 x 10" cells of E . agglomerans were
present per termite gut. This value was loo-fold lower than expected, based on
N2 fixation rates of E. agglomerans in vitro and that of the intact termites.
However. low recoveries of E. agglomerans may be related to the marked
decrease in N2 fixation rates observed when intact termites or their extracted
guts were manipulated for the isolation Of bacteria. It was concluded that the N2-
fixing activity of E. agglomerans may be important to the N economy of C.

formosanus.

 

In 1973 Breznak et al. (7) demonstrated N,
fixation in termites by using the acetylene
(0.1-1,) reduction assay. These workers showed
that the activity was associated with the ter-
mite gut, could be modulated by the amount of
combined N in the diet of the termites and
could be abolished by feeding the insects anti-
bacterial drugs, indicating that termite gut
bacteria mediated N, fixation. It was sug-
gested that N,-i’ixing bacteria or their meta-
bolic products might be important as an N
source for some termites since the food (wood) of
the insects is relatively low in combined N.
Benemann (4) also observed N2 fixation in ter-
mites and reported variations in Cally-reducing
activity between different groups of Kalotermes
minor. Recently, Breznak (6) found that 02H,-
reducing activity of Coptotermes formosanus
can vary over ZOO-fold, with high rates being
exhibited by young, growing larvae. In fact, it
was estimated that the amount of N, fixed by
young larvae could, over the period of a year,
allow the termites to double their N content if
the fixation rate remained constant.

These findings and the suggestion that bacte-
rial N, fixation might be important to some
termites during their development (4, 6, 7)
prompted a search for the organisms involved.

'Jourmal article no. 7760 from the Michigan Agricul-
tural Experiment Station.

392

The results of such an endeavor constitute the
substance of the present paper.

(A portion of this work was presented at the
76th Annual Meeting of the American Society
for Microbiology, 2-7 May 1976. Atlantic City,
N.J.)

MATERIALS AND METHODS

Termites. Formosan subterranean termites. C.
formosanus Shiraki, were collected in the vicinity of
Lake Charles, La., and maintained in the labora-
tory in the form of termite-infested wood. Externally
undifferentiated larvae beyond the second instar
(i.e., worker termites; 23) were used. The average
fresh weight of workers was 2 mg. The same group of
termites was used for all experiments.

Isolation of Enlcrobaclcr agglomerans. A suc-
cessful enrichment medium consisted of equal por-
tions of the following two solutions, which were
sterilized separately. Solution 1 contained (milli-
grams per 100 ml of distilled water): glucose, ribose,
glycerol. and mannitol, 800 each; MgSO, - 711,0, 100;
NaCl, 2; FeSO.-7H,0. 3; Na,MoO.-2H,O. l;
CaCl,-2H,O. 13; sodium thioglycolate, 100; and res-
azurin, 0.2. Solution 2 contained 1% (vol/vol) of a
vitamin solution (34) in 0.1 M potassium phosphate
buffer. pH 7.0. The medium was prepared under N,-
CO, (95:5). using strict anaerobic techniques (22).
and was contained in anaerobic culture tubes (Hun-
gate'type; Bellco Glass, Inc., Vineland, NJ.) at a
volume of 5 nil/tube. The final pH of the medium
was 6.9. Solid medium was prepared by incorporat-
ing 1.5% lonsgar no. 2 (Colab Laboratories. Inc.,

106

VOL. 33, 1977

Chicago Hts, Ill.) and omitting thioglycolate.

Workers of C. formosanus were held in a sterile
petri dish and irradiated for 15 s with a germicidal
lamp (30 W; General Electric Co., Schenectady,
NY.) positioned 45 cm from the insects. Termites
were then transferred to an anaerobic glove box
(Coy Manufacturing Co., Ann Arbor, Mich.; 2). and
their guts were removed by using sterile forceps.
Ten guts were placed in a small tissue homogenizer
containing 2.0 ml of sterile enrichment medium and
homogenized for 1 to 2 min. A 0.1-ml amount of the
homogenate (i.e., 0.5 gut equivalents) was inocu-
lated into 4.9 ml of enrichment medium, and this
constituted the 10" dilution. Thereafter. serial 10-
fold dilutions were made up to 10“. Syringe tech-
niques (26) were used for all dilutions. Tubes were
then secured horizontally in a gyratory water bath
shaker (model G76; New Brunswick Scientific Co.,
New Brunswick, NJ.) operating at 176 rpm. The
incubation temperature was 30°C.

The highest dilutions developing visible turbidity
were tested for C,I-I,-reducing activity (see below).
Positive cultures were again serially diluted to 10‘“
and reincubated. Isolates were obtained from the
second 10" dilution tubes by streaking roll tubes
(22) and were considered to be pure cultures after
four successive passages in roll tubes.

Other bacterial strains. Known strains of E. ag-
glomerans, used for comparative purposes, were

. CDC 811-74 and CDC 156-74. The former strain be-

longed to biogroup G2, whereas the latter belonged
to biogroup 2 (14). Both strains were obtained from
W. H. Ewing, Center for Disease Control. Atlanta.
Ga.

Growth studies. Basal medium GSV was used in
all growth studies. Its composition was similar to
that of the enrichment medium, except ribose, glyc-
erol, mannitol, thioglycolate, and resazurin were
omitted, and the glucose concentration was in-
creased to 1%.

For anaerobic cultivation, GSV medium was pre-
pared by boiling individual solutions for 10 min
under a stream of 0,-free Ar or N,, dispensing them
into anaerobically maintained vessels, and combin-
ing appropriate solutions after heat sterilization.
Usually. tubes containing 5 ml of medium were
used. However, some experiments also made use of
l-liter Erlenmeyer flasks containing 300 ml of me-
dium and equipped with serum stoppered sampling
ports and gas (N,) inlet and outlet tubes. For aerobic
cultivation, media were prepared without anaerobic
precautions and dispensed in 50-ml amounts into
300-ml-capacity Nephelo culture flasks equipped
with a side arm (12 by 130 mm; Bellco Glass, Inc.,
Vineland, N.J.). Most cultures were incubated with
shaking as described above. Anaerobic cultures in 1»
liter Erlenmeyer flasks were vigorously stirred with
a magnetic stirrer driving a stirring bar which was
included in the culture vessel. The final pH of all
media used in growth studies was 7.0 t 0.1.

Media contained in tubes or Nephelo flasks were
inoculated with 0.05% (vol/vol) of a broth culture
containing between 1 x 10" and 4 x 10'l cells/ml. A
1% inoculum was used for media in l-liter Erlen-
meyer flasks. Growth was measured turbidimetri-

N,-FIXING BACTERIA FROM TERMITE GUTS

393

cally and by direct cell counts and in some cases also
by viable cell counts and protein determinations.
Turbidimetric measurements were made by using a
Bausch & Lomb Spectronic 20 colorimeter operating
at 660 nm. A Petroff-Hausser counting chamber was
used for making direct cell counts. Viable cell counts
were determined by diluting samples of culture fluid
in nutrient broth (Difco Laboratories, Detroit,
Mich.) and spreading 0.1 ml of appropriate dilutions
(in triplicate) on plates of nutrient agar (Difco).
Plates were then incubated aerobically at 30°C, and
colonies were enumerated after 24 h. Protein was as-
sayed with the Folin phenol reagent after treatment
of samples with NaOH (19). Crystalline bovine se-
rum albumin (Sigma Chemical Co., St. Louis, Mo.)
was used as a protein standard.

Analyses of fermentation products. Cells were
grown under Ar in GSV medium in which 0.107
NILCI (sterilized separately as a 10‘} solution) was
incorporated. Gaseous products were determined by
use of a fermentation "train" (27). At the end of the
fermentation, the medium was clarified (27) and the
clarified fermentation liquor (CFL) was assayed di.
rectly for glucose (31). acetoin and diacetyl (33), and
glycerol (28).

Ethanol was qualitatively identified and quanti-
fied in neutral volatile distillates (27) of CFL by gas
chromatography. A Varian model 2440 gas chroma-
tograph. equipped with an H, flame ionization de-
tector, was used. The column was stainless steel
(0.125 inch by 5 feet (ca. 0.32 by 152.4 cm]) contain-
ing Porapak Q (80/100 mesh; Waters Associates Inc.
Milford, Mass). N, was the carrier gas (30 ml/min).
and temperatures were as follows: column, 170°C;
injector and detector, 205°C each.

Neutral ether extracts (27) of CFL were assayed
for 2.3-butanediol by gas chromatography as de
scribed above, except that temperatures were: col-
umn. 200°C; injector and detector, 250°C each. The
limit of detection using this system was 0.5 nmol of
butanediol per ml of neutral ether extract.

Organic acids were extracted from acidified CFL
with ether (27) and were qualitatively identified and
quantified by gas chromatography. A Varian model
1420 gas chromatograph, equipped with a thermal
conductivity detector, was used. The column was
stainless steel (0.125 inch by 6 feet [ca. 0.32 by 182.9
cm]) and contained 15‘} SP 1220-17: HnPO. on Chro-
mosorb W AW (100/120 mesh; Supelco, Inc., Bele-
fonte, Pa.). Helium was the carrier gas (25 ml/min).
and temperatures were: column, 135°C; injector and
detector, 165°C each. Volatile fatty acids were ap-
plied to the column as ether solutions. Nonvolatile
acids were first methylated and extracted into
CHCl, (20) prior to application. Colorimetric assays
were also used for the estimation of pyruvate and
oxaloacetate (18) and lactate (3).

Biochemical tests. Most biochemical tests were
performed by the methods of Edwards and Ewing
(13). Sugar fermentation reactions were evaluated
by using phenol red broth base (Difco) in which 1‘7.
of the test sugar was incorporated. Sugars were
filter sterilized separately as 10'} solutions. except
esculin, which was sterilized with the broth base.

KCN sensitivity was determined as described by

107

394 POTRIKUS AND BREZNAK

Holding and Collee (21). Nitrate reduction was
tested as described by Lennette et al. (25). using
nitrate reduction broth.

Determination of G+C content in DNA. The mole
percent guanine plus cytosine (G+C) in deoxyribo-
nucleic acid (DNA) was determined by M. Mandel,
using the buoyant density method (30).

C,H, reduction assays. The ability of samples to
reduce C,H, to C,H. was taken as presumptive evi-
dence for N, fixation (29). C,H. was measured by
using flame ionization gas chromatography (29) and
stainless-steel column (0.125 inch by 5 feet) contain-
ing Porapak N (80/100 mesh; Waters Associates,
Inc). N, was the carrier gas (30 ml/min), and the
column temperature was 62°C.

C,H,-reducing activity in enrichment cultures
was measured by injecting 0.3 cm;1 of C,H, into tubes
through the rubber septa and then reincubating
cultures for 24 h before assaying for the presence of
C,H. in the gas phase. For growth studies. assays
were performed by injecting culture samples into
serum stoppered, Ar-filled vials at a volume of 2.5 to
5% of the vial capacity. C,H, was then introduced in
an amount yielding 0.05 atm of C,H, in the gas
phase. and samples were incubated for 1 h at 30°C on
a reciprocating shaker operating at 88 oscillations/
min. The reaction was terminated with 0.4 ml of 50%
trichloroacetic acid’ml of sample, and the head
space gas was immediately assayed for C,H,. 0,
inhibition studies were performed in a similar man-
ner. except that C,H, and varying amounts of 0,
were injected into Ar-filled vials before the introduc-
tion of cells. 0, used in these studies was found to be
free from detectable‘H, and CO by gas chromato-
graphic analysis (24).

C,H, reduction assays of intact termites or their
extracted guts were performed in a manner similar
to that described previously (7).

Other experimental procedures. The Gram stain
was performed with Kopeloff reagents (20). Motility
of cells was determined by direct observation of wet
mounts of broth cultures, using phasecontrast mi-
croscopy.

For electron microscopy, cells were negatively
stained (5) and examined with a Philips EM 300
electron microscope.

The pH of termite gut macerates was measured
by placing 10 guts in 0.05 ml of deionized, glass-
distilled water on a sheet of dental wax. Guts were
then minced with a scalpel. and the pH of the result-
ing macerate was measured with a combination pH
electrode (model 6020; Ingold Electrodes, Inc., Lex-
ington, Mass.) adapted to a Radiometer pH meter
(model 26; The London Co., Copenhagen, Denmark).

RESULTS

Isolation of bacteria. In three independent
experiments the highest dilutions of gut ho-
mogenate yielding growth in primary enrich-
ments were the 10'3 dilutions (two experi-
ments) and the 10" dilution (one experiment).
These also showed C,H,-reducing activity and
were tentatively judged to contain N,-fixing

Am. ENViaosi. Micaosim.

cells. From 10-3 dilution tubes two isolates of
putative N,-fixing bacteria were obtained, and
these (strains C-1 and C-2) were used for fur-
ther study. For approximating the numbers of
N,-fixing bacteria per termite gut. the three
series of diluted gut homogenates were consid-
ered as one series (three tubes per dilution)
from a single termite population. Consultation
of a three-tube most-probablemumber table
(10) indicated that approximately 2 x 10'-' N,-
fixing bacteria were present per termite gut.

Numerous additional attempts were made to
obtain growth and C,H, reduction activity in
primary dilution tubes greater than 10’“. Aero-
bic enrichments were unsuccessful in this re-
gard. as were the following modifications of the
medium: (i) omission of thioglycolate or substi-
tution of this compound with dithiothreitol
(0.025%) or glutathione (0.05%); (ii) supplemen-
tation of the enrichment medium with 0.05%
cholesterol plus 0.3% sodium succinate plus
0.03% yeast extract, 0.3% sodium succinate plus
0.005% casein hydrolysate, 0.1% ethanol plus
0.4% sodium fumarate plus 0.1% Na,SO,, or
0.001% yeast extract alone; (iii) substitution of
sugars in the enrichment medium with 0.5%
sodium pyruvate plus 0.5% sodium formate
plus 0.03% glutamine, or 1% sodium lactate
plus 0.03% glutamine plus 0.005% serine. Al-
though some of these modified media (i.e.,
those containing amino acids or yeast extract)
yielded visible growth at 10" dilution, none
showed C,H,-reducing activity. Subcultures
from such tubes, in media containing lesser
amounts of combined nitrogen, either yielded
no growth or sparse growth, but never 02H,
reduction.

General characteristics of isolates. Isolates
were gram-negative, nonsporeforming, faculta-
tively anaerobic rods, 0.5 by 1.0 pm in size.
Cells were motile, and electron microscopy re-
vealed the presence of peritrichous flagella.
Cells grown aerobically on nutrient agar
formed colonies that were 2 to 4 mm in diame-
ter, round, and white to cream color.

Biochemical reactions of isolates are shown
in Table 1. In addition, sugar fermentation
tests indicated that both strains fermented ara-
binose, dulcitol, esculin, glucose, inositol. mal-
tose, mannitol, rhamnose, salicin, sucrose, tre-
halose, and xylose. Only strain C-1 fermented
lactose. Neither strain fermented adonitol. ery-
thritol, raffmose, or sorbitol. These data indi-
cated that isolates were strains of E . agglomer-
ans (14). In accord with the biogroup designa-
tions of Ewing and Fife (14), strain C-1 was
assigned to aerogenic biogroup G2 (indole nega-
tive, Voges-Proskauer negative), whereas

108

V0]... 33, 1977

Tau 1. Biochemical reactions of strains C-1 and
C-2

 

Reaction' of strain:
Test or substrate

 

C-1 C-2

 

TSl‘
Slant Acid Acid
Base Acid Acid
Gas + -
H,S — -

Urease - -

Indole

Methyl red (37°C)

Voges-Proskauer (37°C)

Citrate (Simmons)

KCN

Lysine decarboxylase

Ornithine decarboxylase

Arginine dihydrolase

Jordan’s tartrate

Nitrate to nitrite

Oxidation-fermentation

Oxidase

l+llil
l

i~ri+i
('114»!

 

' +, Positive reaction; 3, weak positive reaction;
-, negative reaction; I", fermentation.
' TSI, triple sugar iron agar (Difco).

strain C-2 was assigned to anaerogenic bio-
group 2 (nitrate positive, indole negative,
Voges-Proskauer negative).

Fermentation products. To buttress the con-
tention that termite isolates were strains of E.
agglomerans and to learn more about the phys-
iology of the isolates, the products of glucose
fermentation by growing cells were compared
with those of a known strain of E . agglomerans
(Table 2). Major fermentation products of all
strains were 00,, H,, lactate, acetate, ethanol,
and succinate. Small amounts of formate, pyru-
vate, oxaloacetate, acetoin. and diacetyl were
also formed. Whereas strains C-1 and CDC 811-
74 formed significant amounts of glycerol,
strain C-2 formed only trace amounts of this
compound. None of the strains formed 2.3-bu-
tanediol. Carbon recoveries, based on fermen-
tation products alone, were roughly 90%. Oxi-
dation-reduction indexes were close to 1.0, a
value that would be expected from a fermenta-
tion of glucose.

G+C content of DNA. The moles percent
G+C values in the DNAs of termite isolates
were 53.1 (strain 01) and 52.6 (strain 02).
Values for known strains of E. agglomerans
were 56.1 (strain CDC 811-74) and 57.1 (strain
CDC 15674). Although the values for termite
isolates were slightly lower than those of the
two known strains tested, more comprehensive
analyses of the DNA base composition of E.
agglomerans (i.e., the Herbicola-Lathyri bacte-

N,-FIXING BACTERIA FROM TERMITE GUTS

395

ria) have revealed values ranging from 52.6 to
57.7 mol% G+C (32). Consequently, we will
refer to termite isolates as E. agglomerans
strains C-1 and C-2 for the remainder of the
paper.

Growth studies and C,H, reduction tests.
Growth studies were used to verify the N,-
fixing ability ofisolates and to characterize this
activity. GSV basal medium, containing glu-
cose, N-free salts, and vitamins, was used. Pep-
tone, NHiCl, or KN03, when added to GSV.
served as an N source for cells growing either
aerobically, or anaerobically under Ar (Table
3). Cell yields under these conditions ranged
from 4 x 10" to 1 x 10" cells/ml, with higher
yields being obtained aerobically. No growth
occurred in unsupplemented GSV medium aer-
obically or under Ar (Table 3). However, when

TABLE 2. Fermentation products of termite isolates
and E. agglomerans CDC 811-74

 

mmol/100 mmol of glucose fermented

 

 

Product . . E. agglomerans
Strain C-l Strain C-2 CDC 811-74

CO, 103.2 107.3 101.4
H, 103.0 103.5 101.4
Lactate 26.7 51.7 32.7
Acetate 47.0 32.7 27.3
Ethanol 67.9 71.2 65.9
Succinate 14.0 10 4 15.0
Formate 25.0 7 (‘I 5.7
Glycerol 15.5 0.3 15.1
Pyruvate 4.7 6.0 3.1
Oxaloacetate 0.6 3. 9 2.1
Acetoin 0.3 0.2 0.5
Diacetyl 0.5 0.4 0.4
2,3-Butane- 0.0 0.0 0.0

diol
Carbon re- 93.4 92.6 86.4

covered (%)
O-R index" 1.0 1.0 0.9

 

" O-R, Oxidation-reduction.

TABLE 3. Specific growth rates“ of E. agglomerans
strains isolated from termites

 

Doublings x h‘I at initial gas phase

 

 

 

Addition'
to GSV Strain C-1 Strain C-2

basal me-

dium Air IN? 100% Air 100% 100%

N, Ar N, Ar

Peptone 1.56 0.47 0.61 1.64 0.59 0.59
NH.CI 1.17 0.30 0.40 1.14 0.38 0.47
KNO, 0.68 0.34 0.44 0.47 0.35 0.43
None 0.00 0.16 0.00 0.00 0.13 0.00

 

" During exponential growth.
‘ Sterilized separately and incorporated at a final
concentration of 0.2%.

396 POTRIKUS AND BREZNAK

100% N, was used as the initial gas phase, cells
exhibited specific growth rates of 0.13 to 0.16
(Table 3) and reached densities of 4 x 10" cells/
ml. Provision of combined N sources to cells
growing under N, increased their specific
growth rates 18 to 45fold (Table 3) and al-
most doubled the cell yields. Although not
shown in Table 3, only cells growing in unsup-
plemented GSV under N, exhibited C,H,-re-
ducing activity.

When strain G2 was grown under N, in GSV
basal medium, an exponential increase in the
optical density of the culture coincided with
exponential increases in protein, viable cell
number. and C,H,-reducing activity (Fig. 1).
The latter reached a maximum of 250 nmol of
C,H, formed per (h x ml) at the late exponen-
tial phase of growth and declined thereafter. It
can be calculated from the data in Fig. 1 that
the greatest C,H,-reducing activity (normal-
ized to viable cell number) occurred at 33 h and
was 104 nmol of C,H. formed per (h x 10" cells).
Formation of C,H, was dependent on the pres-
ence of C,H,. Almost identical results were ob-
tained with strain C-l.

These data indicated that termite isolates fix
N, only under anaerobic conditions, in media

8 v- ' -I 0
. .Calts
44
O
\CZH‘
\
x
\
2T I. t .- 1.0 - s
.z' Protein

V

:‘r-l
‘\

l

O

I

‘

Lou,o mate CELLS/ML

too” «nous 9H. tomato/ma in Int) 93 L06“, '0 "DOTSON/lit

 

 

 

 

A
'0 so so so
nouns

Fla. 1. Growth and CJ-I, reduction exhibited by
E. agglomerans strain C-2. OD, Optical density.

109

Ant. ENVIRON. MICROBIOL.

lacking a major source of combined N, and that
C,H, reduction reflected this activity.

0, inhibition of C,H, reduction. Because
strains C-1 and C-2 could not grow aerobically
in unsupplemented GSV medium (Table 3), it
was suspected that 02 inhibited N, fixation. To
test this, cells growing anaerobically under N,-
fixing conditions (i.e., in unsupplemented GSV
under N,) were placed in vials containing Ar.
C,H,, and various amounts of 0,. As shown in
Fig. 2, as little as 0.01 atm of 0, almost com~
pletely inhibited C,H, reduction by strain C-l.
Virtually identical results were obtained with
strain C-2. These data indicated that E. ag-
glomerans behaves in a manner similar to that
of many other N,-fixing. facultative anaerobes.

Effect of anaerobiosis on C,H, reduction by
intact termites and their extracted guts. In.
tact termites and extracted termite guts were
made anaerobic, and their C,H,-reducing activ-
ity was tested. Some of the anaerobic conditions
simulated those encountered in the procedure
for the isolation of E . agglomerans (Table 4).

Introduction of termites into the anaerobic
glove box rendered them unconscious and re-
sulted in a 100-fold decrease of their normal
C,H,-reducing activity. The activity was not
reacquired by reexposure to air, even though
the animals regained consciousness (Table 4.

 

 

2.0
3
2
5
E
E a
S 1.0
g l'
i
=3
0
i
I
M
0 . a a '
(i) 0.02 0.04 0.06
I",2 (atm)

Fla. 2. 0, inhibition of 0,”, reduction by cells of
E. agglomerans strain C-I . Each point on the curve
represents the average value of three separate deter-
minations.

110

VOL. 33, 1977

N,-FIX1NG BACTERIA FROM TERMITE GUTS

397

TABLE 4. Effect ofanaerobiosis on CJ-I, reduction by C. formosanus and their extracted guts

 

Atmosphere‘ in assay C,Hrreducmg ac-

 

Expt no. Specimen“ Treatment vial ,jvityr
1 IT Untreated (control) Air 0611'1
2 IT Admitted into glove box' and immedi- Ar-H, (90:10) 0.006
ately assayed

3 IT As for expt 2'. but reexposed to air prior Air 0.000
to assay'

4 IT Gassed with Ar for 1 min prior to assay' Ar 0.252

5 IT Gassed with N, for 1 min prior to assay' N, 0.118

6 TG Guts removed in glove box and immedi- Ar-H, (90:10) 0.003
ately assayed

7 TG As for expt 6, but reexposed to air prior Air 0.000
to assay

8 TG Guts removed and assayed aerobically Air 0.000

 

" Groups of 25 specimens were used in each experiment: IT, intact termites; TG. termite guts.
’ Atmospheres also contained C,H, which was injected at zero time (see text).

' Nanomoles of C,H. formed per (hour x 25 specimens).

‘ Equivalent to 12.21 nmol of 0,11. formed per (h x g [fresh weight]).

' Termites were rendered unconscious by this treatment.

’ Termites regained consciousness during the 1-h assay period.

experiments 1, 2, and 3). Although H, (a known
inhibitor of nitrogenase [9]) constituted 10% of
the glove box atmosphere, its presence was not
alone responsible for the loss of C,H,-reducing
activity, since even a brief exposure of intact
termites to pure Ar or N, also had a dramatic
inhibitory effect (Table 4, experiments 4 and 5).
Extracted termite guts showed a similar re-
sponse, even if extracted and assayed aerobi-
cally (Table 4, experiments 6, 7, and 8).

The loss of C,H,-reducing activity by intact
termites did not appear to result from a drastic
change in the pH of their gut contents during
anaerobiosis. Gut macerates of untreated ter~
mites had pH’s of 6.5 and 7.0 (two separate
determinations), whereas those from termites
kept under Ar for 1 h had a pH of 6.9.

DISCUSSION

Although E. agglomerans has been isolated
from a variety of intestinal and extra-intestinal
habitats (15, 16), to our knowledge this is the
first demonstration of E. agglomerans in ter-
mite guts and the first quantitative analysis of
the fermentation products of this species. Un-
like most Klebsielleae, the tribe to which the
genus Enterobacter belongs (11), our strains of
E. agglomerans did not produce 2.3-butanediol
as a major fermentation product. Rather, the
products formed (Table 2) were typical of a
mixed acid fermentation (12), with a significant
production of glycerol by two of the three
strains assayed.

It is noteworthy that Aho et al. (1) also docu-
mented N, fixation by E. agglomerans, using
isolates obtained from decaying white fir trees.

These workers further substantiated the N,-
fixing ability of isolates by demonstration of
”N, incorporation into growing cells. French.
on the other hand, recently reported the isola-
tion of N,-fixing bacteria from Australian ter-
mites, but did not state the identity of the
isolates, their numbers in guts, or their magni-
tudes of N, fixation (17).

Although E. agglomerans was the only N,-
fixing bacterium we isolated from termite guts,
we are reluctant to conclude that it is the major
N, fixer in this habitat, simply because it was
isolated from relatively low dilutions of gut
homogenate. Calculations based on maximum
rates of C,H, reduction by E. agglomerans in
vitro [104 nmol of C,H, formed per (h x 10”
cells)] indicated that a population of 2.3 x 10‘
cells/gut would be necessary to account for the
activity observed with intact termites [0.611
nmol of C,H. formed per (h x 25 termites);
Table 4]. Based on most-probable—number esti-
mates, our isolation attempts yielded 100-fold
fewer cells per gut than the expected value,
even when aerobic isolation procedures were
used or when the enrichment medium was ex-
tensively modified. It is significant, however,
that removal of termite guts, even under anaer-
obic conditions, resulted in a 99 to 100% de-
crease in the C,H,-reducing activity of the prep-
aration (Table 4). A 60 to 100% loss in activity
occurred even when intact termites were made
anaerobic prior to assay (Table 4). Benemann
(4) also observed a 90% loss of C,H,-reducing
activity when K. minor was incubated anaero-
bically. The inhibitory effect of anaerobiosis on
C,H,-reducing activity of termites may have a
bearing on our inability to retrieve E. agglom-

398 POTRIKUS AND BREZNAK

erans in greater numbers. Interestingly, recov-
ery of E. agglomerans in numbers 100«fold
lower than expected paralleled the 100-fold de-
crease in C,H,-reducing activity observed when
the insects were made anaerobic.

Low recoveries of N,-fixing E. agglomerans
may also result from incomplete dispersion of
bacterial aggregates during preparation of gut
homogenates. Electron microscopy of guts of C.
formosanus (8) has revealed the presence of
dense bacterial aggregates that adhere strongly
to the epithelium. If E. agglomerans forms
such aggregates, much more vigorous prepara-
tory procedures may be necessary for their dis-
persal. However. preliminary experiments em-
ploying 0.04% Trition X-lOO or 0.01% Tween 80
in the homogenizing solution or blending gut
homogenates in a microblender assembly were
unsuccessful in this regard. Finally, the pres-
ence of a true microaerophilic N,-fixing bacte-
rium in guts, in numbers greater than those of
E. agglomerans, has not been discounted and is
a possibility currently being investigated in our
laboratory.

In view of these considerations. we prefer to
be conservative at this time and conclude that
E. agglomerans may be important to the N
economy of C. formosanus.

ACKNOWLEDGMENTS

We are grateful to Manley Mandel for performing the
DNA base composition analyses and H. S. Pankratz for help
with electron microscopy. We thank J. M. Tiedje for helpful
comments on the manuscript.

The research was supported by the Michigan Agricul-
tural Experiment Station of Michigan State University and
by National Science Foundation grant BMS75-05850. C. J.
P. is the holder of a National Science Foundation predoc-
toral fellowship.

ADDENDUM IN PROOF

French et al. (J. Gen. Microbiol. 95:202-206, 1976)
recently implicated Citrobacter freundii as a nitro.
gen-fixing agent in guts of Australian termites.
However the number of C. freundii cells per termite
gut was not determined.

LITERATURE CITED

1. Aho. P. E., R. J. Seidler. H. J. Evans, and P. N. Raju.
1974. Distribution. enumeration. and identification of
nitrogen-fixing bacteria associated with decay in liv-
ing white fir trees. Phytopathology 64:1413-1420.

. Aranki, A., S. A. Syed. E. B. Kenney. and R. Freter.
1969. Isolation of anaerobic bacteria from human gin-
give and mouse cecum by means of a simplified glove
box procedure. Appl. Microbiol. 17:568-576.

. Barker. S. B. 1957. Preparation and colorimetric deter-
mination of lactic acid, p. 241-246. In S. P. Colowick
and N. 0. Kaplan (ed.). Methods in enzymology. vol.
3. Academic Press Inc., New York.

. Benemann. J. R. 1973. Nitrogen fixation in termites.
Science 181:164-165.

. Brenner. S.. and R. W. Horne. 1959. A negative stain-

111

10.

11.

12.

13.
14.

15.

16.

17.

18.

19.

20.

21.

23.

24.

25.

Area. ENV1RON. MICROBIOL.

ing method for high resolution electron microscopy of
viruses. Biochim. BiOphys. Acta 34:103-110.

. Breznak. J. A. 1975. Symbiotic relationships between

termites and their intestinal microbiota. p. 559-580
In D. H. Jennings and D. L. Lee (ed.). Symbiosis
(Society for Experimental Biology Symposium Ser.,
no. 29). Cambridge University Press. Cambridge

. Breznak. J. A.. W. J. Brill. J. W. Mertins. and H. C.

Coppel. 1973. Nitrogen fixation in termites. Nature
(London) 244:577-580,

. Breznak. J. A., and H. S. Pankratz. 1977. In situ mor-

phology of the gut microbiota of wood-eating termites
[Reticulitermes flavipes (Kollar) and Coptotermes for-
mosanus Shiraki]. Appl. Environ. Microbiol. 33:406.

. Burris. R. H. 1971. Fixation by free«living micro-orga-

nisms: enzymology, p. 105—160. In J. R. Postgate
(ed.), The chemistry and biochemistry of nitrogen
fixation. Plenum Press. New York.

Collins. C. H.. and P. M. Lyne. 1970. Microbiological
methods, p. 174-190. University Park Press. Balti-
more.

Cowan. S. T. 1974. Family I. Enterobacteriaceae Rahn
1937, p. 290—340. In R. E. Buchanan and N. E. Gib-
bons (ed.). Bergey‘s manual of determinative bacteri-
ology, 8th ed. The Williams and Wilkins Co., Balti-
more.

Doelle. H. W. 1975. Bacterial metabolism. p. 559-692.
Academic Press Inc., New York.

Edwards. P. R., and W. H. Ewing. 1972. Identification
of Enterobacteriaceae, 3rd ed. Burgess Publishing
Co., Minneapolis.

Ewing. W. R., and M. A. Fife. 1972. Enterobacter ag-
glomerans (Beijerinck) comb. nov. (the Herbicola-
Lathyri bacteria). Int. J. Syst. Bacteriol. 22:4-11.

Ewing. W. H.. and M. A. Fife. 1974. Biochemical char-
acterization of Enterobacter agglomerans. Depart-
ment of Health. Education and Welfare publication
no. (CDC) 74-8173. Center for Disease Control. At-
lanta. Ga.

Ewing. W. 11.. and M. A. Fife. 1974. Biochemical char-
acterization of Enterobacter agglomerans. Depart.
ment of Health Education and Welfare publication
no. (CDC) 75-8173. Center for Disease Control, At-
lanta. Ga.

French. J. R. J. 1975. The role of termite hindgut bacte-
ria in wood decomposition. Mater. Org. 10:1-13.

Friedemann. T. E., and G. E. Haugen. 1943. Pyruvic
acid. 11. The determination of keto acids in blood and
urine. J. Biol. Chem. 147:415—442.

Herbert. D.. P. J. Phipps. and R. E. Strange. 1971.
Chemical analysis of microbial cells, p. 209-344. In J.
R. Norris and D. W. Ribbons (ed.). Methods in micro.
biology, vol. 5B. Academic Press Inc., New York.

Holdeman. L. \'.. and W. E. C. Moore. 1972. Anaerobe
laboratory manual. Virginia Polytechnic Institute
and State University, Blacksburg.

Holding. A. J., and J. G. Collee. 1971. Routine biochem-
ical tests. p. 1-32. In J. R. Norris and D. W. Ribbons
(ed.). Methods in microbiology, vol. 6A. Academic
Press Inc., New York.

. Hungate. R. E. 1950. The anaerobic mesophilic cellulo-

lytic bacteria. Bacteriol. Rev. 14:1-49.

King. E. 6.. Jr.. and W. 'I'. Spink. 1974. Laboratory
studies on the biology of the Formosan subterranean
termite with primary emphasis on young colony de-
velopment. Ann. Entomol. Soc. Am. 67:953-958.

Kyryacos. G., and C. E. Boord. 1957. Separation of
hydrogen. oxygen, nitrogen. methane, and carbon
monoxide by gas adsorption chromatography. Anal.
Chem. 29:787-788.

Lennette. E. 11.. E. H. Spaulding. and J. P. Truant.
1974. Manual of clinical microbiology. 2nd ed. Ameri-

112

VOL. 33, 1977

can Society for Microbiology, Washington. DC.

26. Macy. J. M., J. E. Snellen. and R. E. Hungate. 1972.
Use of syringe methods for anaerobiosis. Am. J. Clin.
Nutr. 25:1318—1323.

27. Neish, A C. 1952. Analytical methods for bacterial fer-
mentations. 2nd revision. National Research Council
of Canada. Saskatoon.

28. Pinter. J. K., J. A. Hayashi. and J. A. Watson. 1967.
Enzymatic assay of glycerol. dihydroxyacetone, and
glyceraldehyde. Arch. Biochem. Biophys. 121:404-
414.

29. Postgate. J. R. 1972. The acetylene reduction test for
nitrogen fixation, p. 343-356. In J. R. Norris and D.
W. Ribbons (ed.). Methods in microbiology. vol. 68.
Academic Press Inc., New York.

30. Schildkraut. C. L., J. Marmur. and P. Doty. 1962.

N,-FIXING BACTERIA FROM TERMITE GUTS

399

Determination of the base composition of deoxyribo—
nucleic acid from its buoyant density in CsCl. J. Mol.
Biol. 4:430-443.

31. Spiro. R. G. 1966. Analysis of sugars found in glycopro-
teins, p. 3-26. In E. F. Neufeld and V. Ginsburg (ed.).
Methods in enzymology. vol. 8. Academic Press Inc.,
New York.

32. flan. M. P.. and M. Mandel. 1969. DNA base composi-
tion and taxonomy of phytopathogenic and other
enterobacteria. J. Gen. Microbiol. 56:113-123.

33. Westerfeld. W. W. 1945. A colorimetric determination
ofblood acetoin. J. Biol. Chem. 161:495-502.

34. Wolin. E. A., M. J. Wolin. and R. S. Wolfe. 1963.
Formation of methane by bacterial extracts J. Biol.
Chem. 238:2882-2886.

nICHIan STATE UNIV. LIBRARIES
lliIINWIIHWHIHNIWII‘HHWWIWWIWI
31293105247922