~:12

 

||||lll|UlllllllllHlIHlllIlHUlllllllllngLlllflllJlll

EM“ 3 1293 1052490 LIBRARY

Michigan State
University

 

 

This is to certify that the
thesis entitled
HeterotrOphic Bacteria of Wood-Eating Termites

[Reticulitermes flavipes (Kollar)]

presented by

Joanne Elaine Schultz

has been accepted towards fulfillment
of the requirements for

M. S. degree in Microbiology

 

 

 

0-7639

 

. cl! .lt'EIi... irrul

Elly l1 1,
- nytllr¢\li

 

 

 

moo7ma

HETEROTROPHIC BACTERIA OF WOOD-EATING TERMITES
[RETICULITERMES FLAVIPES (KOLLAR)]

 

By

Joanne Elaine Schultz

A THESIS

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

MASTER OF SCIENCE

Department of Microbiology and Public Health
1978

ABSTRACT

HETEROTROPHIC BACTERIA 0F WOOD-EATING TERMITES
[RETICULITERMES FLAVIPES (KOLLAR)]
By

Joanne Elaine Schultz

To better understand the nature of prokaryotes inhabiting the
gut of xylophagous termites, strict anaerobic culture techniques were
used to quantitate heterotrophic bacteria present in hindguts of

Reticulitermes flavipes. The grand mean number of viable bacteria

 

per hindgut was 0.4 x lo5 (lst instar larvae), l.3 x 105 (3rd instar
larvae), 3.5 x 105 (workers), and 1.5 x 105 (soldiers). Of 344 iso-
lates, 66.3% were homolactic streptococci (mostly S, lggtj§_and §,
cremoris) and l7.4% were bacteroides which were obtained regardless
of the origin of termites, their caste, or length of captivity. Se-
lectedstrains of bacteroides fermented lactate to propionate and
acetate in anaerobic cell suspensions. lg_vitrg cocultures of S,

lactis and Bacteroides sp. revealed that lactate formed by S, lactis

 

was fermented as an energy source by Bacteroides sp. to propionate,

 

acetate, and 602. Results indicate that not only are streptococci
and bacteroides relatively stable components of the hindgut micro-
biota, but also that cross-feeding of lactate between these micro-
organisms may constitute one aspect of the overall hindgut fermenta-

tion in the termite.

DEDICATION

To my parents
who have given me encouragement and support

in all endeavors, past and present.

ii

ACKNOWLEDGMENTS

I wish to express my sincere appreciation to Dr. John A.
Breznak for his guidance, encouragement, and patience throughout
the course of this study. I am especially grateful to him for finan-
cial support and the opportunity to work and learn in his laboratory.

I also wish to thank members of my guidance committee, Dr.
C. A. Reddy, Dr. M. J. Klug, and Dr. J. R. Miller, for helpful dis-
cussions. Technical assistance received from Mr. Clifford Sporn is
gratefully acknowledged.

Deep appreciation is expressed to two very special friends
and colleagues, Ms. Catherine Potrikus and Mr. Michael Spellman,
for their moral support and understanding during this study. Their
friendship has been a source of strength for me that I shall never

forget.

TABLE OF CONTENTS

 

 

 

Page
LIST OF TABLES ......................... v
LIST OF FIGURES ........................ vi
INTRODUCTION .......................... l
LITERATURE REVIEW ....................... 3
Biology of Termites . . ' ................. 3
Symbiotic Relationships between Termites and Their
Intestinal Microorganisms ............... 6
Lower Termites .................... 6
Protozoan Symbionts ............... 6
Bacterial Symbionts ............... 9
Higher Termites ................... l4
LITERATURE CITED ........................ 18
ARTICLE 1: HETEROTROPHIC BACTERIA PRESENT IN HINDGUTS OF
WOOD-EATING TERMITES [RETICULITERMES FLAVIPES
(KOLLAR)] ..................... 25
ARTICLE 2: CROSS-FEEDING OF LACTATE BETWEEN STREPTOCOCCUS
LACTIS AND BACTEROIDES SP. ISOLATED FROM TERMITE
HINDGUTS ...................... 33

iv

LIST OF TABLES

Table Page
Article 1

1. Summary of bacteria isolated from guts of R, flavipes
workers ......................... 28

2. General properties of bacteria isolated from guts of R,
flavipes workers .................... 29

3. Bacteria isolated from guts of freshly collected larvae,
soldiers, and workers of R, flavipes ..........

4. Physiological reactions of Streptococcus isolates . . . . 30

Article 2

l. Fermentation of glucose-UL-14C by S, lactis le in
monoculture and in coculture with Bacteroides sp. JN20 . 42

 

2. Effect of lactate on growth and product formation by
Bacteroides awzo in monoculture ............. 47

 

LIST OF FIGURES

Figure Page
Article 2

1. Surface colonies of Streptococcus lactis strain JWl
(large colonies) and Bacteroides sp. strain JNZO
(smaller colonies) on a lOO mm diameter plate of
supplemented BHI medium ................. 39

 

 

2. Growth, substrate utilization, and product formation
by a coculture of Streptococcus lactis strain le and
Bacteroides sp. strain szo ............... 44

 

 

vi

INTRODUCTION

Termites are social insects which ingest cellulosic materials
for nutrition. It has long been known that protozoa in the hindgut
of evolutionarily "lower" termites are responsible for the digestion
of cellulose into useable nutrients for the termite. However, the
hindgut also contains an abundance of morphologically diverse bacteria
which appear to be an integral part of the gut ecosystem (10). while
bacteria in termite hindguts have been implicated in nitrogen fixa-
tion (9) and methane production (8), very little is known about their
other roles in the hindgut. Uncertainty about the importance of bac-
teria in termite metabolism arises largely from the paucity of data
concerning the numbers and species composition of the bacterial com-
ponent, and their biochemical capabilities.

It was felt that a knowledge of the major types of heterotrophic
bacteria in the termite gut would provide a basis for studying the
nutritional aspects of these bacteria in the hindgut environment.
Therefore, a study was initiated with the following objectives: l) to
isolate under nonselective conditions, characterize, and quantify the
major groups of heterotrophic bacteria found in the wood-eating ter-

mite, Reticulitermes flavipes; 2) to compare the bacterial component

 

of different castes and developmental stages of R, flavipes; 3) to de-
termine the influence of laboratory captivity and the origin of R,

f1§gjpg§_on bacterial types; and 4) to investigate, with jn_vitro

techniques, nutritional aspects of the major groups of bacteria and

possible interactions among these bacteria in the termite hindgut.

LITERATURE REVIEW

Biology of Termites

 

Termites are insects belonging to the order Isoptera which in-
cludes five living families: Mastotermitidae, Kalotermitidae, Hodo-
termitidae, Rhinotermitidae, and Termitidae (90). The first four
families are referred to as "lower" termites, since they are phylo-
genetically more primitive; members of the Termitidae are referred
to as "higher" termites. Lower termites all possess intestinal
symbiotic protozoa on which they depend for digestion of cellulose.
The Termitidae, which contain three-fourths of all known species,
harbor protozoa which are neither numerous nor cellulolytic (35).

The termite colony is usually produced by a pair of reproduc-
tives and consists of several castes, which are morphologically and
functionally different (soldiers, alates, and neotenics), in addition
to other individuals in various stages of development (larvae, winged
nymphs, and workers) (24). Newly hatched larvae proceed through de-
velopmental stages (instars) in which the termite undergoes ecdysis.
Depending on the needs of the colony, larvae can differentiate into
reproductives or soldiers, or remain undifferentiated (9l). Workers
are most abundant in the termite colony and consequently, dominate
all colony activities except reproduction and defense. However, in
many species workers do not constitute a terminal caste and are there-

fore difficult to define. For example, workers of the genus

4

Reticulitermes are merely defined as externally undifferentiated

 

larvae past the third instar (24). These forms, like the lst to 3rd
instar larvae, retain the potential to differentiate into soldiers or
reproductives.

The diet of termites consists of plant material, either living,
dead, or in various stages of decay by fungi. The majority of the
lower termites feed on wood, while the Termitidae prefer a more varied
diet of grass, humus, dead wood, or plant debris (57). Many termites
prefer wood decayed by fungi and benefit from the fungal degradation
of lignin and other wood constituents (77). Termite species, however,
react differently to fungi species. While some termites might favor
a particular fungus, others may be indifferent or repelled by it (3).

Workers obtain and digest crude nutrients to feed themselves
and dependent reproductives, soldiers, and larvae. In lower termites,
dependent castes solicit stomodeal food (a mixture of saliva and re-
gurgitated raw food) or proctodeal food from the workers (67). Proc-
todeal food consists of liquid excretions from rectal pouches and is
excreted in response to tactile stimuli (57). Higher termites do not
practice proctodeal feeding, although coprophagy has been reported (65).

The intestinal tract of a worker termite consists of a foregut,
midgut, and hindgut. An esophagus, crop, and gizzard comprise the
foregut, while the midgut is a simple tube of uniform diameter. Di-
gestive enzymes, which degrade easily hydrolyzable compounds, are
secreted primarily in the midgut (67). In some Termitidae, the midgut
is elongated on one of the faces of the intestinal tube, resulting in
a "mixed segment" (67). An enteric valve separates the midgut from

the hindgut which, in turn, consists of five regions including a

well-developed, bulbous paunch. It is the paunch region which contains
the bulk of microorganisms (bacteria and protozoa) responsible for
most of the digestion in termites. The paunch may thus be likened to
a fermentation chamber superficially similar to the rumen of cattle.
The paunch contents displays a pH close to neutrality (31, 49, 67, 75),
and products of digestion are absorbed in this region since the entericl
valve prevents refluxing of paunch contents to the midgut (67).
Cellulose and hemicellulose, the major polysaccharides of plant
material, undergo extensive degradation on passage through the termite
gut. Oshima (68) first demonstrated cellulose digestion in termites

from analyses of the constituents of wood and feces of Coptotermes

 

 

formosanus. Hungate (44) compared wood and fecal pellets of Zooterm-
gp§j§_and found that 84% of the cellulose was utilized. Seifert and
Becker (78), using five different woods and four species of termites,
found that 74 to 99% of the initial cellulose was degraded. Other
studies have found hemicelluloses to be substantially attacked (25,

37, 47). Esenther and Kirk (25) fed aspen sapwood to Reticulitermes

 

flavipes and found 84% of the major hemicellulose, xylan, and 77% of
the mannan-type hemicellulose was utilized on passage through the
termite.

The extent of lignin degradation by termites appears minor.
Seifert and Becker (78) obtained values for lignin degradation as

high as 77% with Reticulitermes santonensis. However, this estimate

 

appears to be high because of uncontrolled coprophagy. By contrast,

studies with R, flavipes (25) and Microcerotermes edentatus (47)

 

revealed no net degradation of lignin after passage through termites,

although a slight loss in the methoxyl content of the lignin was

6

detected. Other studies either indicate that lignin degradation does
not occur in termites (20, 37) or provide data of questionable signif-
icance (29).

Symbiotic Relationships between Termites
'and Their Intestinal Microorganisms

 

 

As defined by deBary (2), symbiosis is an intimate and rela-
tively stable association of two dissimilar organisms. Embraced by
this definition is the relationship between termites and their intes-
tinal microorganisms. The symbiosis between lower termites and their
hindgut protozoa can best be described as mutualistic (l2). Oxymonad,
trichomonad, and hypermastigote flagellates (35) degrade cellulose
into useable nutrients for the termite, while the termite provides
the protozoa with food and lodging. By contrast, the association be-
tween termites and their intestinal bacteria is less well-understood,
although several lines of recent evidence suggest that they, too,
may be important to the termites' nutrition and vitality. Protozoan

and bacterial symbionts will be discussed individually below.

Lower Termites

 

Protozoan Symbionts

Cleveland (12, l3, l4) recognized the ability of termites to
thrive on sound wood or cellulose, and he showed that hindgut proto-
zoa were indispensable for cellulose digestion and survival of the

insects. Reticulitermes flavipes could be defaunated (i.e. rid of

 

their protozoa) by incubation at elevated temperatures, or under
hyperbaric 02, or by starvation, without apparent damage to the in-

sect per_§e_(l5). However, if such termites were fed a normal diet of

wood, death occurred within 10-20 days. Indefinite survival of de-

faunated termites could be achieved by refaunation, or by feeding the
termites fungus-digested wood or cellulose (12). These data indicated
that the cellulose-digesting activity of hindgut protozoa was critical

to termite survival. Selective defaunation of Zootermopsis (15, 16),

 

however indicated that not all protozoa were of equal value to termite

survival. The presence of Leidyopsis or Trichonympha could keep the

 

 

termite alive indefinitely, whereas Streblomastix could not. Other

 

workers have used defaunation techniques to determine the importance
of protozoa in termite survival (22, 79, 80, 81, 92) or cellulose
catabolism and lipid synthesis (62, 64).

Cellulase activity was first demonstrated by Trager (88) in

guts of Reticulitermes flavipes, Zootermopsis angusticollis and in the

 

 

xlephagous cockroach, Cryptocercus punctulatus. When extracts of the

 

guts were incubated with purified cellulose, an increase in reducing
sugars was observed. Analysis of extracts of foregut, midgut, and
hindgut showed that cellulase activity was confined to the hindgut
region (19, 37, 76, 88). That the protozoa were the source of the
enzyme was shown by the lack of cellulase activity in both defaunated
termites (37) and cockroaches (19, 88). In addition, cellulase activ-

ity has been demonstrated in cultures of Trichomonas termqpsidis, a

 

flagellate of Zootermopsis (88, 92). While most data indicate that

 

the protozoa are the sole source of cellulase, Yokoe (93) found that
about 20% of the original cellulase activity was maintained in de-

faunated Leucotermes speratus. He suggested that the termite produced

 

the cellulase, however no mention was made as to whether the bacterial

component was also affected by the defaunation techniques.

8

Cleveland (13) postulated that the protozoa benefitted the
termite by degrading cellulose to cellobiose and then to glucose.

By using an extract of the protozoan I, termopsidis, Trager (88)

 

showed that glucose was a product of cellulose breakdown. It was
thus supposed that the protozoa excreted glucose which was absorbed
by the termite for carbon and energy. However, it was not until
Hungate's jg_vjtrg_experiments with protozoa from Zootermopsis, that
the products of cellulose digestion were quantified (38, 40). He
found that 70-75% of the cellulose carbon could be accounted for in
CO2 and acetic acid. H2 was also produced, but did not appear to be
utilized by the termite. Only small amounts of lactic acid were de-
tected. Hungate (38, 40, 42) suggested that the protozoa obtained
energy through anaerobic fermentation of cellulose, liberating C02,
H2, and organic acids (primarily acetic). Furthermore, his respiro-
metric data with live termites suggested that acetate was subsequently
absorbed by the termite and oxidized for energy. Examination of gut

contents revealed the presence of acetate in hindguts of Zootermopsis

 

(40) and pr0pionate (in addition to acetate) in hindguts of unidenti-
fied lower termites (11). These data suggest that defaunated termites
should be able to survive if fed acetate. However, such attempts have
been unsuccessful (21, 42), despite the permeability of the termite
gut to this compound (40).

More extensive knowledge of the nutrition and metabolism of
termite flagellates is limited, primarily because axenic culture con-
ditions have not been developed. Trager (89) succeeded in culturing

Trichomonas termopsidis from Zootermopsis anaerobically on a cellulose-

 

 

based medium for over 3 years, although his cultures also contained

a noncellulolytic bacillus. Other attempts to cultivate cellulolytic

protozoa have been less successful (34, 44, 89). Trichonympha

 

sphaerica from Zootermopsis survived as long as 6 weeks on an anaerobic

 

cellulose medium, but growth was not maintained in subcultures (89).

Bacterial Symbionts
Besides protozoa, bacteria are also abundant in the intestinal

tract of lower termites. Direct microscopic counts by Breznak et_al,

6

(9) revealed an average of 3 x 10 bacteria per gut in workers of

 

Coptotermes formosanus. Most of the bacteria have been found to re-
side in the hindgut (9, 52, 53). Electron microscopy studies by
Breznak and Pankratz (10) revealed that the paunch epithelium of R,

flavipes and C, formosanus is densely colonized by morphologically

 

diverse bacteria, many of which possess holdfast elements that secure
them to the epithelium and other bacterial cells. Aggregation of bac-
teria near indentations on the epithelial surface, possibly involved
in ion absorption, suggest that bacteria could be nutritionally or
biochemically important to the termite. In addition, Breznak and

Pankratz (10) found that the midgut of R, flavipes and C, formosanus

 

contained endospore-forming bacteria of a single morphologicaltype.
Despite the abundance of bacteria in termite guts, few attempts

have been made to isolate them with nonselective media and to identify

them. Krasil'nikov and Satdykov (53) isolated gut bacteria from two

species of Anacanthotermes by plating diluted intestinal contents on

 

eleven different media. They found that Enterobacter aerogenes,

 

Enterobacter cloacae, and Streptococcus faecalis were always isolated,
5

 

 

ranging from about 4 x 10 to l x 109 viable bacteria per gut depending

on the caste of termite or its stage of development. In addition,

lO

Escherichia coli, Pseudomonas liquifaciens, and Bacillus subtilis were

 

 

also isolated from intestines, but less frequently (53). Mauldin et_al,

(63) estimated that approximately 3 x 105 viable bacteria were present

 

per hindgut of C, formosanus by using a tryptone-glucose-yeast extract
medium, but he did not identify the isolates. Eutick e; 31, (26) re-
cently isolated gut bacteria from 9 different species of Australian
termites, by using plates of reinforced Clostridial medium. Quantita-
tion and identification of the bacteria revealed that each termite
species possessed one major bacterial type in concentrations of approx-

imately 107 cells per m1 of termite gut. Streptococcus was isolated

 

as the major bacterium from families Mastotermitidae and Kalotermitidae,

while Enterobacter was found in 4 species of Rhinotermitidae. Strep-

 

tococcus was also isolated from 3 species of Rhinotermitidae, but in

quantities 10- to lOO-fold less than that of Enterobacter. Although

 

isolation plates were incubated under both aerobic and anaerobic con-
ditions, no strict anaerobes were detected (26).

Most previous studies of termite gut bacteria have focussed
on their possible involvement in cellulose digestion. Accordingly,
many investigators have employed enrichment cultures as a means of
detecting such fbrms. Cleveland (12, 13) was unsuccessful in more
than 50 attempts to detect cellulolytic bacteria in guts of R,
flavipes. Likewise, Dickman (23) and Hungate (36) used anaerobic .

and aerobic media and failed to detect cellulolytic bacteria in gut

 

contents.of Zootermopsis. Eutick et 31. (26) found no evidence of
cellulolytic bacteria in 9 different species of termites, although
cellulolytic bacteria were readily isolated from soil.

Several investigators have obtained limited success with -

ll

enrichments for cellulolytic bacteria. Beckwith and Rose (4) inocu-
lated gut contents of 7 species of lower termites into cellulose
media.. While no cellulose decomposition was evident under anaerobic
conditions, 9 of 64 aerobic cultures indicated degradation of the
polysaccharide. The time necessary for degradation varied from 10
days to 3 months, and microscopic examination revealed gram negative
rods and some micrococci. Mannesmann (58) also used enrichment cul-
tures to detect cellulolytic bacteria in gut contents of Reticuli-

termes virginicus and Q, formosanus. Although some destruction of

 

 

filter paper was evident after 10 weeks, bacteria grew in all cultures
including many control tubes containing no cellulose. Other workers
(28, 54, 55, 82, 83, 84) also claimed the isolation of cellulolytic
bacteria from guts of lower termites. Unfortunately, a lack of quan-
titative methodology has made it difficult to ascertain the importance

of such isolates in cellulose digestion in situ. To date, there is

 

no convincing evidence that cellulolytic bacteria are important to
the nutrition of lower termites which possess cellulolytic protozoa.
Attention has also focussed on the importance of bacteria in
the nitrogen economy of termites. This is not surprising inasmuch
as wood, the food of many termites, is low in combined N (i.e. 0.03-
0.05% N w/w; 39). .Cleveland (14) postulated that atmospheric nitro-
gen (N2) was fixed by termites. He analyzed air samples from tubes

containing live Zootermopsis, but could not detect any change in the

 

amount of N2 present. Likewise, nitrogen balances fbr colonies of

Zootermopsis showed no net gain in combined nitrogen indicating that

 

N2 was not fixed (39, 4l). By contrast, a specific assay for N2 fixation,

acetylene reduction, has been used to confirm N2 fixation in termites

12

(5, 9, 30). Breznak (8) estimated that the amount of N2 fixed by
young larvae could allow such fbrms to double their protoplasmic N

in the period of a year if their fixation rate remained constant.

In addition, Breznak et_al, (9) fbund that gut bacteria were respon-
sible, since elimination of bacteria corresponded to a loss in the
termite's ability to fix N2. Whereas bacteria have long been impli-
cated in N2 fixation (33, 69, 70, 85, 86), the isolation and charac-
terization of such bacteria had not been done. Recently, however,
Potrikus and Breznak (72) isolated Nz-fixing Enterobacter agglomerans

from hindguts of Q, formosanus and implicated this species in N2

 

fixation in situ. The recoveries of g, agglomerans from guts was

lOO-fold lower than expected, based on N2 fixation rates of g,

 

agglomerans in_vitro and that of the intact termites. However, E,

agglomerans was the only Nz-fixer isolated under a variety of condi-

 

tions. French et_al, (30) isolated NZ-fixing Citrobacter freundii

from guts of Q, lacteus and Mastotermes darwiniensis, however the

 

bacteria were not quantitated. Likewise, no quantitative data are

available for Nz-fixing Enterobacter from several species of lower

 

termites (26).

Bacteria have also been implicated in the recycling of nitroge-
nous excretory compounds in the termite colony. Leach and Granovsky
(56) first postulated that microorganisms in the hindgut of termites
could utilize urates and other nitrogenous waste products of termites.
Isolation of uricolytic bacteria from R, flavipes has revealed numbers
as high as 6 x 104 cells per hindgut (73). Isolates were identified

as strains of Streptococcus, Bacteroides, and Citrobacter and were

 

 

found to degrade uric acid under anaerobic conditions.

l3

Termites have been shown to emit methane, which appears to be
produced by gut bacteria (8). When 5, flavipes larvae were fed a diet
of nest wood, 27-73 nmoles of methane were emitted per hour per gram
of fresh weight. Bacteria seem to be responsible for the production
of methane since feeding larvae antibacterial drugs abolished methane'
emission and detectable bacteria, but not protozoa (8).

Many bacteria in the hindguts of termites are associated with
protozoa, either as residents within the cytoplasm or attached to the
protozoan surface. A variety of bacteria have been reported to be
endosymbiotic (6, 45, 46, 69). Bloodgood (6) found rod-shaped or
fusiform-shaped bacteria, which showed no signs of degeneration, en-

closed.in membrane vesicles within the cytoplasm of Pyrsonympha and

 

Trichonympha of R, flavipes. 'He suggested that essential nutrients

 

were transferred from the protozoan cytoplasm to the prokaryotes.
The prokaryotes, on the other hand, may confer on protozoa the abil-
ity to degrade cellulose or fix atmospheric nitrogen (6, 69, 70).
Hungate (37) attempted to culture such intracellular bacteria, but
without success.

Many bacteria, including spirochetes, are associated with
protozoa by attachment to the protozoan surface. Bloodgood (6)
fbund that spirochetes. associated with the surface membrane of
Pyrsonympha of R. flavipes, provided the specialized structure for
attachment to the protozoa. On the other hand, Urinympha and Barbula-

 

nympha of _C_. punctulatus provided attachment structures promoting
the association of rod-shaped bacteria. Little is known of the roles
of the attached bacteria, although the associations are believed to

be advantagous to one or both partners (6). The attachment of

14

spirochetes to the flagellated Mixotrichia paradoxa is the only case

 

where a role for the prokaryote has been established. Cleveland and
Brimstone (18) found that the host protozoan derives its motility
from synchronous movement of its adherent spirochetes. Margulis a
21, (60) hypothesized that adherent spirochetes, originally selected
to confer motility on the protozoan, evolved into eukaryotic flagella.
Consistent with this notion is the recent evidence for microtubules

in certain spirochetes (and a gliding bacterium) from hindguts of
subterranean and drywood termites (60). No comparable tubules were
seen in any of the other bacteria from the termite hindgut.

The biochemical capabilities of spirochetes in the termite
hindgut are not known, primarily because the spirochetes have never
been isolated in pure culture (7). Cleveland (17) eliminated spiro-
chetes from the guts of lower termites by feeding them filter paper
moistened with a solution of 5% acid fuchsin. He discovered that the
removal of Spirochetes did not affect the vitality of the termite.
Eutick gt_al, (27) attempted to eliminate spirochetes from Q, lacteus

and the higher termite, Nasutitermes exitiosus, but found that none

 

of the flora, including the spirochetes, were affected by acid fuchsin.
All attempts that eliminated the spirochetes in C, lacteus also re-
moved the protozoa. However, selective removal of spirochetes in N,

exitiosus significantly reduced the termite's life span.

Higher Termites

 

Higher termites do not possess the cellulolytic protozoa fOund
in lower termites, although they do have a substantial bacterial pop-

ulation in their guts. The majority of bacteria are located in the

15

.paunch region, although the mixed segment, when present, usually con-
tains bacteria of a single morphological type (32, 50). Potts and
Hewitt (74) described a variety of paunch bacteria, including spirilla,

small rods, and cocci in Trinervitermes trinervoides. They also found,

 

in agreement with other workers (32, 50), no bacteria in the foregut
or midgut and "pure cultures" of rods in mixed segments. Microscopic

examination of guts of Macrotermes subhyalinus revealed an abundance

 

of cocci and small gram positive rods in the paunch (l). Pochon et_

21, (71) fbund the paunch of Sphaerotermes sphaerothorax to contain

 

' cocci, rods, fusiforms, and spiral-shaped bacteria, many of which
were similar to fbrms seen in the rumen of cattle.
Several genera of bacteria have been isolated from guts of

higher termites. Strains of Pseudomonas (59) and cellulolytic

 

Achromobacter (28) were isolated from guts of Nasutitermes exitiosus,

 

 

but were not quantitated. Eutick et a1, (26) isolated Staphylococcus
7

 

from three species of Nasutitermes in concentrations of about 10 cells

 

per ml of termite gut. Hungate (43) isolated a cellulolytic actino-

mycete, Micromonospora pr0pionici, from Amitermes minimus in quanti-

 

 

 

ties of about 500 colony-forming-units per gut. However, he concluded
the bacterium to be of limited importance in the digestion of cellu-
lose in situ.

It is still speculative whether bacteria play a role in the
digestion of cellulose. Misra and Rangathan (66) suggested that the

cellulase and cellobiase found in the hindgut of Termes obesus were

 

derived from bacteria, but they presented no conclusive evidence for
this. Enrichment cultures for cellulolytic bacteria from guts of S,

sphaerothorax indicated growth of cocci, primarily diplococci, rods,

 

l6

and some vibrio-shaped bacteria (71). Analysis of the mixed culture
fluid revealed the presence of acetate and butyrate, but no lactate
or succinate. Tracey and Youatt (87) confirmed the presence of
cellulase in guts of N, exitiosus, but did not determine its source.
Kovoor (51) found cellulase activity in all intestinal regions of

Microcerotermes edentatus and attributed it to the bacteria as well

 

as synthesis by the termite. Other workers found cellulase activity
localized in the midgut, a region essentially devoid of bacteria
(1, 61, 74). Potts and Hewitt (74) found approximately 70% of the

cellulase activity in the midgut of I, trinervoides, with 40% of

 

this activity associated with the midgut wall. They concluded that
cellulase was synthesized by the termite, although about 50% of the
cellobiase was found in the paunch and could have been of bacterial
origin.

Cellulolytic activities have recently been studied in fungus-

 

growing termites, Macrotermes. Abo-Khatwa (l) fbund Cx cellulase
(active against soluble derivatives, such as carboxymethylcellulose)
and cellobiase equally distributed between the midgut and paunch and
concluded that bacteria in the hindgut may contribute to digestion by
providing Cx enzymes to the termite. Martin and Martin (61) impli-
cated the midgut as the major site for these enzymes and concluded
from analyses of midgut tissue and salivary glands that one source

of Cx cellulase and cellobiase may be the termite itself. Abo-Khatwa
(l) and Martin and Martin (61) found the vast majority of C1 cellulase
(active against crystalline cellulose) located in midgut contents.

Fungi grown by Macrotermes sp. in nest material on "fungus combs"

 

were analyzed for the presence of cellulolytic enzymes. Conidiophores

17

of the fungus Termitomyces, which are ingested by the termites, were

 

found to contain high levels of C1 cellulase and Cx cellulase (l, 61).
If termites were starved or fed combs without fungus nodules, C1 cellu-
lase activity significantly decreased. It was suggested that the in-
gested condidiophores contributed C1 cellulase to aid the termite in
the initial stages of cellulose digestion.

Regardless of the origin of the cellulase, an active fermenta-
tion appears to take place in the paunch region of higher termites.
Kovoor (48) detected acetic, propionic, and butyric acids in hindguts

of Microcerotermes endentatus. Mannesmann (59) quantitated acids in

 

the hindgut fluid of Nasutitermes nigriceps and found 4.7 mg of

 

acetate per gram of gut fluid. Lactate levels approached only 0.5
mg per gram of hindgut fluid. The presence of acetate, propionate,
butyrate, and only small amounts of lactate in guts of higher termites
indicate an overall fermentation similar to that occurring in hindguts

of lower termites.

10,

11.

LITERATURE CITED

Abo-Khatwa, N. 1978. Cellulase of fungus-growing termites:
A new hypothesis on its origin. Experientia 343559-560.

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

Becker, G. 1971. Physiological influences on wood-destroying
insects of wood compounds and substances produced by micro-
organisms. Wood Sci. Tech. §,236-246. '

Beckwith, T. D., and E. J. Rose. 1929. Cellulose digestion by
organisms from the termite gut. Proc. Soc. Expt. Biol. Med.
27:4-6.

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

Bloodgood, R. A., and T. P. Fitzharris. 1976. Specific associa-
tions of prokaryotes with symbiotic flagellate protozoa from
the hindgut of the termite Reticulitermes and the wood-eating
roach Cryptocercus. Cytobios 113103-122.

 

 

Breznak, J. A. 1973. Biology of nonpathogenic, host-associated
spirochetes. CRC Crit. Rev. Microbiol. 2,457-489.

Breznak, J. A. 1975. Symbiotic relationships between termites and
their intestinal microbiota, p. 559-580. In D. H. Jennings
and D. L. Lee (eds.), S mbiosis (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 244,577-580.

Breznak, J. A., and H. S. Pankratz. 1977. In situ morphology of
the gut microbiota of wood-eating termites [Reticulitermes
flavipes (Kollar) and Coptotermes formosanus Shiraki] Appl.
Environ. Microbiol. 335406-426.

 

 

Brown, 8., and R. Smith. 1954. The relationship of the concen-
tration of acetic acid in normal and defaunated termites.
Mendelian 25,19-20.

18

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

19

Cleveland, L. R. 1923. Symbiosis between termites and their
intestinal protozoa. Proc. Natl. Acad. Sci. USA 23424-428.

Cleveland, L. R. 1924. The physiological and symbiotic rela-
tionships between the intestinal protozoa of termites and
their host, with special reference to Reticulitermes flavipes
Kollar. Biol. Bull. 463178-227.

 

Cleveland, L. R. 1925. The ability of termites to live perhaps
indefinitely on a diet of pure cellulose. Biol. Bull. 483
289-293.

Cleveland, L. R. 1925. The effects of oxygenation and starva-
tion on the symbiosis between the termite, Termopsis, and
its intestinal flagellates. Biol. Bul. 483309-327.

Cleveland, L. R. 1926. Symbiosis among animals with special
reference to termites and their intestinal flagellates. 0.
Rev. Biol. 1351-60.

Cleveland, L. R. 1928. Further observations and experiments
on the symbiosis between termites and their intestinal pro-
toza. Biol. Bull. 543231-237.

Cleveland, L. R., and A. V. Grimstone. 1964. The fine structure
of the flagellate Mixotrichia paradoxa and its associated
microorganisms. Proc. R. Soc. Lond. B 1593668-686.

 

Cleveland, L. R., S. R. Hall, E. P. Sanders, and J. Collier. 1934.
The wood-feeding roach Cryptocercus, its protozoa, and the
symbiosis between protozoa and roach. Mem. Am. Acad. Arts
Sci. 11:184-342.

 

Cohen, W. E. 1933. An analysis of termite (Eutermes exitiosus)
mound material. J. Council Sci. Ind. Res. Aust. 63166-169.

 

Cook, S. F. 1943. Nonsymbiotic utilization of carbohydrates by
the termite, Zootermopsis angusticollis. Physiol. Zool.1§3
123-128.

 

Cook, S. F., and R. E. Smith. 1942. Metabolic relations in the
termite-protozoa symbiosis: temperature effects. J. Cell.
Comp. Physiol. 193211-219.

Dickman, A. 1931. Studies on the intestinal flora of termites
with reference to their ability to digest cellulose. Biol.
Bull. 61385-92.

Esenther, G. R. 1969. Termites in Wisconsin. Ann. Entomol.
Soc. Am. 6231274-1284.

Esenther, G. R., and T. K. Kirk. 1974. Catabolism of aspen
sapwood in Reticulitermes flavipes. Ann. Entomol. Soc. Am.
613989-991.

 

26.

27.

28.

29.

31.

32.

33.

35.

36.

37.

38.

39.

20

Eutick, M. L., R. W. O'Brien, and M. Slaytor. 1978. Bacteria
from the gut of Australian termites. Appl. Environ. Microbial.
333823-828.

Eutick, M. L., P. Veivers, R. W. O'Brien, and M. Slaytor. 1978.
Dependence of the higher termite, Nasutitermes exitiosus and
the lower termite, Captatermes lacteus on their gut flora.

J. Insect Physiol. 343363-368.

 

 

French, J. R. J. 1975. The role of termite hindgut bacteria in
wood decomposition. Mater. Org. 1931-13.

French, J. R. J., and D. E. Bland. 1975. Lignin degradation in
the termites Coptotermes lacteus and Nasutitermes exitiosus.
Mater. Org. 193281-288.

 

 

French, J. R. J., G. L. Turner, and J. F. Bradbury. 1976. Nitra-
gen fixation by bacteria from the hindgut of termites. J. Gen.
Microbial. 333202-206.

Grassé, P. P., and Ch. Nairot. 1945. La transmission des
flagellés symbiotiques et 1es aliments des termites. Bull.
Biol. Fr. Belg. 133273-292.

Grassé, P. P., and Ch. Nairot. 1959. L'évalutian de la symbiase
chez 1es Isaptéres. Experientia 133365-372.

Green, R. A., and E. L. Breazeale. 1937. Bacteria and the nitro-
gen metabolism of termites. J. Bacteriol. 33395-96.

Gutierrez, J. 1956. The metabolism of cellulose-digesting,
symbiotic flagellates of the genus Trichonympha from the
termite Zootermgpsis. J. Protozoal. 3339-42.

 

 

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

 

Hungate, R. E. .1936. Studies on the nutrition of Zootermopsis.
I. The role of bacteria and molds in cellulose decomposition.
Zent. Bakterial. Parasitenkd. Infektianskr. Abt. 2 343240-249.

 

Hungate, R. E. 1938. Studies on the nutrition of Zootermopsis.
II. The relative importance of the termite and the protozoa
in wood digestion. Ecology 1331-25.

 

Hungate, R. E. 1939. Experiments on the nutrition of Zooter-
mo sis. III. The anaerobic carbohydrate dissimilatian by the
intestinal protozoa. Ecology 393230-245.

 

Hungate, R. E. 1941. Experiments on the nitrogen economy of
termites. Ann. Entomol. Soc. Am. 343467-489.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

21

Hungate, R. E. 1943. Quantitative analyses on the cellulose
fermentation by termite protozoa. Ann. Entomol. Soc. Am.
333730-739.

Hungate, R. E. 1944. Termite growth and nitrogen utilization
in laboratory cultures. Texas Acad. Sci. Proc. Trans. 313
1-7.

Hungate, R. E. 1946. The symbiotic utilization of cellulose.
J. Elisha Mitchell Sci. Soc. 3339-24.

Hungate, R. E. 1946. Studies on cellulose fermentation. II.
An anaerobic cellulose-decomposing actinamycete, Micramana-
spora pr0pionici, N. sp. 51:51-56.

 

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

 

 

Jirovec, O. 1931. Natizen fiber parasitische Protisten. II.
Symbiose van Bakterien und Trichonympha serbica (Geargevitch).
Zent. Bakterial. Parasitendk. Abt. 1 Originale 1333184-186.

 

Kirby, H. Jr. 1941. Organisms living on and in protozoa, p. 1009-
1113. In G. N. Calkins and F. M. Summers (eds.), Protozoa in
Biological Research. Columbia University Press, New York.

 

 

Kovoor, J. 1964. Modifications chimiques pravaquées par un
Termitidé (Microcerotermes edentatus, Was.) dan du baid de
peuplier sain ou partiellement dégradé par des champignans.
Bul1. Biol. Fr. Be1g. 233491-510.

 

Kovoor, J. 1967. Présence d'acides gras valatils dans la
panse d'un Termite supérieur (Microcerotermes edentatus Was.,
Amitermitinae). C. R. Hebd. Seances AcadLTSCi. Ser. 0
@43486-488.

 

Kovoor, J. 1967. Le pH intestinal d'un Termite supérieur
(Microcerotermes edentatus, Was., Amitermitinae). Insectes
Saciaux, Paris 143157-160.

 

Kovoor, J. 1968. L'intestin d'un Termite supérieur (Micro-
cerotermes edentatus, Was., Amitermitinae). Histaphysialagie
et flare bactérienne symbiotique. Bull. Biol. Fr. Belg. 1933
46-84.

 

 

Kovoor, J. 1970. Presence d'enzymes cellulalytiques dans
l'intestin d'un Termite supérieur, Microcerotermes edentatus
(Was.). Ann. Sci. Nat. Zaal. Biol. Anim. 13365-71.

 

Krasil'nikav, N. A., and Sh. I. Satdykav. 1969. Estimation of
the total bacteria in the intestines of termites. Microbi-
ology 333289-292.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

22

Krasil'nikav, N. A., and Sh. I. Satdykav. 1970. Bacteria of
termites' intestines. Microbiology 333562-564.

Krelinava, 0., V. Jirku, and J. Skoda. 1977. The cellulolytic
activity of some intestinal bacteria of termites. Int.
Biadeteriar. Bull. 13381-87.

Krelinava, 0., V. Jirku, and J. Skoda. 1977. Cellulolytic
activity of bacterial symbiotes of wood-destroying insects.
Folia Microbiol. 333134-136.

Leach, J. G., and A. A. Granovsky. 1938. Nitrogen in the nu-
trition of termites. Science 31366-67.

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

 

Mannesmann, R. 1972. A comparison between cellulolytic bacteria
of the termites Coptotermes formosanus Shiraki and Reticuli-
termes virginicus (Banks.). Int. Biadeteriar. Bull. 33104-111.

 

 

Mannesmann, R. 1977. Untersuchungen des Glucose-Staffwechsels
van Gfirkammerbakterien der Termite Nasutitermes nigriceps
(Haldemann). Z. angew. Entomol. 3331-10.

 

Margulis, L., L. To, and 0. Chase. 1978. Microtubules in pro-
karyates. Science 33931118-1124.

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

sis: The rale of acquired digestive enzymes. Science
19931453-1455.

 

Mauldin, J. K. 1977. Cellulose catabolism and lipid synthesis
by normally and abnormally faunated termites, Reticulitermes
flavipes. Insect Biochem. 1327-31.

 

Mauldin, J. K., N. M. Rich, and D. W. Cook. 1978. Amino acid
synthesis from 14C-acetate by normally and abnormally
faunated termites, Coptotermes formosanus. Insect Biochem.
2:209-217.

 

Mauldin, J. K., R. V. Symthe, and C. C. Baxter. 1972. Cellulose
catabolism and lipid synthesis by the subterranean termite,
Coptotermes formosanus. Insect Biochem. 33209-217.

 

McMahan, E. A. 1969. Feeding relationships and radioisotope
techniques, p. 387-406. In K. Krishna and F. M. Weesner
(eds.), Biology of Termites, vol. 1. Academic Press, New York.

 

Misra, J. N. and V. Ranganathan. 1954. Digestion of cellulose
by the mound building termite, Termes (Cyclatermes) obesus
(Rambur). Proc. Indian Acad. Sci. 333100-113.

 

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

23

Nairot. Ch., and C. Noirot-Timothee. 1969. The digestive system,
p. 49-88. In K. Krishna and F. M. Weesner (ed.), Bialagy of
Termites, vol. 1. Academic Press, New York.

 

Oshima, M. 1919. Formosan termites and methods of preventing
their damage. Philipp. J. Sci. 133319-383.

Pierantoni, U. 1936. La simbiasi fisialagica nei Termitidi
xilafagi e nei laro flagellati intestinali. Arch. 2001.
Ital. 333135-173.

Pierantoni, U. 1937. Osservaziani sulla simbiasi nei Termitidi
xilafagi e nei laro flagellati intestinali II. Defaunaziane
per digiuna. Arch. Zaol. Ital. 343193-207.

Pochon, J., H. deBarjac, and A. Roche. 1959. Recherches sur
la digestion de la cellulose chez 1e termite Sphaerotermes
sphaerothorax. Ann. Inst. Pasteur, Paris 333352-355.

 

 

Potrikus, C. J. and J. A. Breznak. 1977. Nitrogen-fixing Entero-
bacter agg]amerans isolated from guts of wood-eating termites.
Appl. Environ. Microbial. 333392-399.

 

 

 

Potrikus, C. J., and J. A. Breznak. 1978. Uric Acid and uri-
colytic gut bacteria in termites. Abstr. Annu. Mtg. ASM
1:27.

Potts, R. C., and P. H. Hewitt. 1973. The distribution of in-
testinal bacteria and cellulase activity in the harvester
termite Trinervitermes trinervoides (Nasutitermitinae). In-
sectes Sacianx, Paris 333215-220.

 

Randall, M., and Doady, T. C. 1934. Hydrogen-ion concentration
in the termite intestine, p. 94-99. In C. A. Kafaid (ed.),
Termites and Termite Control, 2nd ed. Univ. of California
Press, BerkeTey.

 

Retief, L. W., and P. H. Hewitt. 1973. Digestive B-glucasidases
of the harvester termite Hadatermes mossambicus: properties
and distribution. J. Insect Physiol. 1331837-1847.

 

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

 

Seifert, K., and G. Becker. 1965. Der chemische Abbau van
Laub- und Nadelhalzarten durch verschiedene Termiten.
Holzfarschung 133105-111.

Symthe, R. V. 1972. Feeding and survival at constant temperatures
by normally and abnormally faunated RetiQQJitermes virginicus
(Isoptera: Rhinotermitidae). Ann. Entomol. Soc. Am. 33:756-
757.

 

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

24

Symthe, R. V., and J. K. Mauldin. 1972. Soldier differentiation,
survival, and wood consumption by normally and abnormally
faunated workers of the formasan termite, Coptotermes forma-
sanus. Ann. Entomol. Soc. Am. 3331001-1004.

 

Symthe, R. V., and L. H. Williams. 1972. Feeding and survival
of two subterranean termite species at constant temperatures.
Ann. Entomol. Soc. Am. 333226-228.

Tetrault, P. A., and W. L. Weis. 1937. Cellulose decomposition
by a bacterial culture from the intestinal tract of termites.
J. Bacteriol. 33395.

Thayer, D. W. 1976. Facultative wood-digesting bacteria from
the hind-gut of the termite Reticulitermes hesperus. J. Gen.
Microbial. 333287-296.

 

Thayer, D. W. 1978. Carboxymethylcellulase produced by faculta-
tive bacteria from hind-gut of the termite Reticulitermes
hesperus. J. Gen. Microbial. 133313-18.

 

Tdth, L. 1948. Nitrogen-binding by Kalotermes flavicallis
(Isoptera) and its symbionts. Hung. Acta Biol. 1322-29.

 

T6th, L. 1952. The role of nitrogen-active microorganisms in
the nitrogen metabolism of insects. Tijdshrift Entomol. 333
43-62.

Tracey, M. V., and G. Youatt. 1958. Cellulase and chitinase
in two species of Australian termites. Enzymalogia 13370-72.

Trager, W. 1932. A cellulase from the symbiotic intestinal
flagellates of termites and of the roach, Cryptocercus punctu-
latus. Biochem. J. 3331762-1771.

 

Trager, W. 1934. The cultivation of a cellulose-digesting
flagellate, Trichomonas termopsidis, and of certain other
termite protozoa. Biol. Bull. 333182-190.

 

Weesner, F. M. 1960. Evolution and biology of the termites.
Annu. Rev. Entomol. 33153-170.

Weesner, F. M. 1965. The Termites of the Unites States, a
Handbook. The National Pest Control Association, New Jersey.

 

Yamin, M. A. 1978. Studies on Trichomitopsis termopsidis
(Cleveland): Cellulolytic activity of this flagellate and
prolonged survival of defaunated termites after reinfection.
J. Protozaol . 33:22A.

 

Yokoe, Y. 1964. Cellulase activity in the termite, Leucotermes
speratus, with new evidence in support of a cellulase produced
by the termite itself. Sci. Pap. Coll. Gen. Educ. Univ.

Takyo 143115-120.

 

ARTICLE 1

HETEROTROPHIC BACTERIA PRESENT IN HINDGUTS OF
WOOD-EATING TERMITES [RETICULITERMES

 

FLAVIPES (KOLLAR)]

By J. E. Schultz and John A. Breznak

Reprinted from Appl. Environ. Microbial. 333930-936.

25

26

APPLIED AND ENVIRONMENTAL Micnoaiowov. May 1978, p. 930—936

(1199-2240/78/(X135-093030200/0
Copyright © 1978 American Society for Microbiology

Vol. 35, No. 5
Printed in U. SA.

Heterotrophic Bacteria Present in Hindguts of Wood- Eating

Termites [Reticulitermes flavipes (Kollar)]‘l'

J. E. SCHULTZ AND JOHN A. BREZNAK'

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

Received for publication 21 November 1977

Strict anaerobic culture techniques were used to quantitate heterotrophic
bacteria present in hindguts of Reticulitermes flavipes. The grand mean number
of viable cells per hindgut was 0.4 x 105 (first-instar larvae), 1.3 X 101" (third-instar
larvae), 3.5 x 105 (workers), and 1.5 X 105 (soldiers). Of a total of 344 isolates,
66.3% were streptococci that were always obtained regardless of the origin of
termites, their developmental stage or caste, or their length of captivity. Most of
the remaining isolates were strains of Bacteroides and Enterobacteriaceae. A
small percentage were strains of Lactobacillus, Fusobacterium, and unidentified
anaerobic gram-positive rods. Recovery of bacteria from worker hindguts was
13.0% of the direct microscopic count. lsolations performed aerobically failed to
reveal strict aerobes. Attempts to isolate cellulolytic bacteria were uniformly
unsuccessful. Of 145 streptococcal strains isolated from freshly collected termites,
almost all were Streptococcus lactis and S. cremoris. Enterobacteriaceae isolates
from the same termite specimens were indole-positive Citrobacter, citrate-nega-
tive Citrobacter, and Enterobacter cloacae. The possibility of in situ interspecies
lactate transfer, between lactate producers (e.g., streptococci) and lactate fer-

menters (Bacteroides), is discussed.

 

The hindgut of xylophagous lower termites
harbors a microbiota consisting of both protozoa
and bacteria. Although much attention has been
focused on the protozoa (because many are cel-
lulolytic and exist in a mutualistic relationship
with the termites [3, 29]). our understanding of
the bacterial component is meager. However,
available evidence suggests that procaryotes are
also important elements of the gut biota and
important to the vitality of termites. First, they
are abundant (3), and their presence appears
necessary for persistence of the protozoa and for
normal termite nutrition (4, 36). Second, N2
fixation by some of the bacteria has implicated
them in the N economy of termites (4, 18, 33).
Third, many exist in intimate physical associa-
tion with the gut epithelium, a situation presum-
ably reflecting biochemical interaction as well
(5).

Numerous attempts have been made to isolate
cellulolytic bacteria from termites. Hungate (22),
for example, isolated an anaerobic, cellulolytic
actinamycete (Micromonospora propionici)
from guts of Amitermes minimus, but concluded
that the bacterium was of limited importance in
situ. Other such attempts have been unsuccess-
ful (7, ll, 20) or have yielded positive results of
uncertain significance, either because the meth-

1' Journal article no. 8344 from the Michigan Agricultural
Experiment Station.

930

ads were not fully described or because no quan-
titation of bacteria was made (2, 17, 27, 30, 32,
35, 37, 38). By contrast, Krasil’nikov and Sat-
dykov (25) used relatively nonselective media to
quantitate bacteria present in guts of Anacan-
thotermes ahngerianus and A. turkestancius.
They found 1.3 X 10“ to 4.3 x 109 viable cells per
hindgut, depending on the developmental stage
or caste of termite examined. However, these
workers apparently used only aerobic conditions
for bacterial isolation and could have overlooked
a numerically significant population of gut an-
aerobes.

It was felt that a better understanding of the
nature and possible roles of bacteria in the ter-
mite gut required not only a comprehensive and
quantitative approach, but attention to strict
anaerobic methodology as well. Accordingly, the
present study was initiated. In this paper we
report on heterotrophic bacteria present in hind-
guts of Reticulitermes flavipes, the common
eastern subterranean termite.

MATERIALS AND METHODS

Termites. R. flavipes was collected in Janesville,
Wis, and Spring Arbor. Mich., and maintained in the
laboratory as previously described (5), except that
commercial lumber or paper towels were not supplied.
The contents of termite containers were periodically
moistened by spraying the surface of infested wood
with distilled water.

27

V01... 35, 1978

Terminology used to indicate the developmental
stage or caste of termites was that of Esenther (14).

Media. Strict anaerobic techniques (19. 23) were
used for preparation of all media unless indicated
otherwise. Media included: supplemented brain heart
infusion (BHI); chopped meat with meat particles
(CMMP); Sweet E; and peptone-yeast-glucose (PYG).
All were compounded according to Holdeman and
Moore (19) except for PYG, which contained 0.5%
each of peptone, yeast extract, and glucose, 0.05%
cysteine-H01, 10"% resazurin, and inorganic salts
(12). In addition, a medium (MB medium) designed
for isolating saccharolytic rumen bacteria was also
used. The composition of MB medium was described
by Bryant (6) and was used herein with 0.5% glucose,
maltose. cellobiose, or soluble starch or with 1.0% a-
cellulose (Sigma Chemical Co., St. Louis, Mo.) as sole
fermentable carbohydrate. Variations of MB medium
also tried were: (i) omission of the volatile fatty acid
solution; (ii) substitution of volatile fatty acids with
bovine rumen fluid (10% [vol/vol], final concentra-
tion); and, for cellulose-containing media, (iii) incor-
poration of 0.1% Trypticase into variations (i) and (ii).

Solid media contained 1.5 or 2.0% agar. MB media
were prepared as roll tubes (23) with 100% CO: as gas
phase and were contained in anaerobic screw-cap cul-
ture tubes (Hungate type; Bellco Glass Inc., Vineland.
N .J.). Just before use. sterile reducing agent (cysteine-
sulfide solution) was added by syringe (23). All other
media were prereduced and anaerobically sterilized
(19) under 95% N2-5% 002 and were contained in
rubber-stoppered 18- by l42-mm culture tubes
(broths) or in plastic petri plates (solid media) that
were filled while held in an anaerobic glove box (1;
Coy Manufacturing 00., Ann Arbor, Mich.). Media
used under aerobic conditions contained no added
reducing agents or resazurin. The initial pH of media
ranged from 6.7 to 7.2.

Isolation of bacteria. Guts were removed anaer-
obically from groups of 10 termites and were homog-
enized as previously described (33). Serial 10-fold di-
lutions of gut homogenate were made, and 0.1 ml of
each dilution was spread in duplicate on plates of
isolation medium. Broth solutions, used for homoge-
nizing guts and diluting the homogenate, were homol-
ogous to the plating medium but lacked agar. Plates
were then either left directly in the glove box atmos-
phere (90% N2-10% H2) or placed in a GasPak jar
(BBL, Cockeysville, Md.) containing an activated
GasPak H; + CO: generator envelope. Jars were also
retained in the glove box. Incubation was at 23 to 25°C
for at least 7 days. Isolated colonies were then picked
at random and were considered pure cultures after
three successive passages on streak plates.

For isolation of bacteria by the roll tube method
(23) a similar approach was used, except that dilution
of gut homogenates and inoculation of molten agar
media were performed by syringe.

Control experiments used the following source ma-
terials, which were plated on BHI agar: (i) surface
washes of intact termites; (ii) homogenates prepared
from degutted termite bodies; and (iii) homogenates
prepared from termite midguts, portions of which fre-
quently remained attached to extracted hindguts (5).
Midguts were obtained by severing extracted guts just

HETEROTROPHIC BACTERIA IN TERMITE HINDGUTS

931

anterior to the enteric valve.

Enumeration of bacteria. The number of culti-
vable bacteria per gut was calculated from the average
number of colonies present on primary isolation plates
or in roll tubes. The acridine orange-epifluorescence
method (16) was used for determining the total num-
ber of bacteria present in gut homogenates.

Characterization of isolates. Isolates were char-
acterized by standard microbiological methods (8).
Since all were found to grow in BHI, plates of this
medium were used to test for aerobic growth.

Anaerobes were identified to the genus level by
using criteria outlined in the VP1 Anaerobe Labora-
tory Manual (19). Acidic fermentation products were
determined after cells were grown anaerobically for 96
h in BHI broth. Organic acids were extracted from the
spent medium (19) and analyzed by gas chromatogra-
phy (33).

Streptococci were identified to species according to
the criteria of Deibel and Seeley (10). Ammonia for-
mation from arginine was assessed as described by
Deibel (9). Lancefield precipitin tests were performed
by using autoclaved cells according to the bulletin
accompanying the commercial antiserum (Difco Lab-
oratories, Detroit, Mich). Sugar fermentation reac—
tions were evaluated by determining the pH of cultures
72 h after inoculation into a basal medium containing
1.0% (xylose or cellobiose) or 0.5% (xylan) of the test
sugar. The basal medium contained 1.0% tryptone,
0.1% yeast extract, and 4.0% (vol/vol) salts solution
( 19) lacking NaHCO;.. The ability of cells to develop a
final pH at least 1.0 unit lower than that in control
medium (lacking sugar) was interpreted as a positive
result. Xylan (Sigma Chemical Co.) was twice ex-
tracted with boiling distilled water before use in fer-
mentation tests.

Enterobacteriaceae were identified according to es-
tablished criteria and tests (13, 26). In addition, pre-
sumptive tests for the ability of strains to fix N; were
made by using GSV medium and methods previously
described (33).

Lactate fermentation by Bactemidea strains.
Formation of propionate and acetate from lactate by
growing cells was determined by using PY and PYL
media (19).

Lactate fermentation by cell suspensions was as-
sessed by using strains grown anaerobically in BHI
broth. Strict anaerobic techniques were used through-
out. Cells were harvested by centrifugation. resus-
pended in 0.1 M potassium phosphate buffer (pH 7.2)
containing 1 mM dithiothreitol, and added to reaction
mixtures (3.0 ml, final volume) containing (millimolar):
potassium phosphate buffer (pH 7.2), 93.3; dithiothre-
itol, 2.6; and sodium-m.-lactate, 17.0. Reaction mix-
tures were held under 02-free N2 in 5-ml-capacity
serum vials equipped with rubber stoppers. Incubation
was for 2 h at 37°C, after which reactions were ter-
minated by addition of 0.2 ml of 5 N H2804. Acidified
mixtures were centrifuged to obtain cell-free super-
natant fluids, which were then used for assay of organic
acids. Results were corrected for propionate and ace-
tate formation by endogenous cell metabolism. Heat-
treated cell suspensions (100°C for 15 min) served as
controls. Organic acids and cell protein were assayed
as previously described (33).

28

932 SCHULTZ AND BREZNAK

Other bacterial strains. A variety of known bac-
teria. were used as positive and negative controls,
either for the efficacy of the anaerobic methodology
or for physiological/biochemical tests. Such orga-
nisms, and their origin, were: Peptostreptococcu.~: an-
aerobius. Bacteroides fragilis, Bacillus licheniformis,
Escherichia coli, Proteus mirabilts. Enterobacter
cloacae, and Pseudomonas aeruginosa (culture col-
lection. Department of Microbiology and Public
Health. Michigan State University); Streptococcus
cremoris ATCC 19257, S. lactis A'l‘CC 19435, and S.
faecalis ATCC 19433 (American Type Culture Collec-
tion, Rockville, Md.).

RESULTS

Bacteria isolated from guts of R. flavipes
workers. A summary of isolates obtained with
BHI medium is given in Table 1. No marked
differences were observed for termites collected
from sites 237 miles (ca. 381 km) apart, whether
isolations were performed immediately after col-
lection or 17 months thereafter. The grand mean
number of cultivable bacteria per gut was 3.5 I
2.8 x 10" (n = 13). The mean number obtained
by using Ng/Hg incubation atmosphere was not
significantly different at the 01111.": level from that
observed with GasPak incubation (Student's t
test; Table l, footnote c). Moreover, the distri—
bution of physiological types retrieved was vir-
tually identical under either incubation atmos-
phere (data not shown). Streptococci were al-
ways obtained and were clearly the predominant
cultivable organisms, accounting for 66.3"}.
(138/208) of all isolates. Most of the remainder
were represented by strains of Bacteroides and
Enterobacteriaceae, which together accounted
for 27.9% of the total and which were isolated

APPL. luvnum. Mlcuomou
almost routinely. Strains of Lactobacillus, Fu-
sobacterium, and unidentified anaerobic rods
constituted a minority of isolates.

When surface washes of live termites or de-
gutted termite bodies were used as inocula, col-
ony counts with BHI medium were 2 x 10"-fold
less, indicating that the isolates described in
Table l were not extraintestinal contaminants.
Moreover, lOO-fold fewer colonies were obtained
from midgut homogenates, indicating that iso-
lates were virtually all derived from hindguts.

The ability to isolate strict anaerobes from
guts served as an internal control for the meth-
odology used. However, an additional control
consisted of known strains of B. fragilis and P.
anaerohius that were manipulated in a manner
simulating the isolation procedure with guts.
Both strains formed isolated colonies after 4
days of incubation in either the Nz/Hz or GasPak
atmosphere.

During the initial phase of this study, a variety
of isolation media were screened for their suita-
bility. It was found that r:covery of bacteria on
CMMP, Sweet E, or PYG medium was not
significantly different from that obtained with
BHI, either in terms of the number of viable
cells per gut or the nature of the isolates. By
contrast, recovery of bacteria with the various
MB media (containing soluble sugar) was only
""1” to 1.4. that obtained with BHI; however, the
isolates were not characterized. Ultimately BHI
medium was chosen for routine use because it
permitted rapid and luxurious colony develop-
ment, and it was readily prepared.

Aerobic isolation attempts revealed an aver-
age of 1.4 x 10” viable cells per gut (one experi-

TABLE 1. Summary of but-term isolated from guts of R. flavipes workers"

 

 

Length Viable No. of
Location and date of ('01- ot cap- bacteria/ isolates
lection tivity gut (X charac- Strepto-
(mo)" lt) ")‘ terized cuccus
Janesville. Wis.
June 1977 0 2.1 44 28
June 1976 2 2.0 33 19
June 1976 9 3.0 16 13
June 1975 9 8.6 38 17
June 1973 17 5.8 23 15
Spring Arbor. Mich.
May 1977 0 1.4 30 27
July 1976 8 1.6 24 I9

Generic affiliation of isolates

 

 

 

Bacteroides I" ‘ Uniden-
"‘— ”Enteric Lm'lo- u. "- tified an-
...) bm-tc- .
. . , type bacillus . aerobic
(uroup I bump 2 Hum _
rods
14 I I 0 0 0
8 l 4 0 0 l
l 0 2 0 0 0
4 0 9 ti 0 2
0 3 2 l 2 0
2 l 0 0
0 0 (l 0 0

 

" All isolates were obtained by using BHI medium.

" Zero months refers to termites used within 24 h of collection.

' Average of two experiments (one using Nz/Hg isolation atmosphere. the other using Gasl’ak) for each period
of captivity, except for 9-months captivity of June, 1076, Janesville collection (one experiment; Nz/Hz atmos-
phere). Grand mean. 3.5 :t 2.8 x 10" (n = 13). Mean value with Nz/Hg atmosphere, 3.1 :1: 2.0 x 10'" (n s 7); with

Gasl’ak atmosphere. 4.0 :t 3.7 x 10" (n = 6).
“ Assigned to the family Enterobacter-(acme.

29

VOL. 35, 1978

ment). Of 41 random isolates examined. all were
facultative anaerobes, but they were not char-
acterized further. All attempts to isolate cellu-
lolytic bacteria were unsuccessful, although con-
trol experiments (using sheep rumen contents as
inoculum) yielded such organisms at high dilu-
tion.

Direct microscopic counts revealed an average
of 2.7 X 10" bacteria per gut ( n = 4). Comparison
of this datum with the grand mean number of
viable cells per gut indicated a recovery of about
13%. Gram-stained preparations of gut homoge-
nates revealed numerous gram-negative rods
and spiral-shaped cells along with gram-positive
cocci. Gram-positive rods were rarely observed,
and most protozoa were disrupted by the ho-
mogenization procedure.

General properties of bacteria isolated
from R. flavipes workers. Results are pre-
sented in Table 2. All isolates were acidogenic
and developed a final pH of about 4.5 in BHI
broth as a result of organic acid production.

Streptococci occurred as short chains of coc-
coid (i.e., football-shaped) cells. All were homo-
lactic fermenters, forming no detectable gas in
BHI broth. All but one strain were facultative
anaerobes. Cells were consistently variable with
respect to the Gram reaction, even when 8-h
broth cultures were examined. In addition, all
strains fermented xylose and cellobiose; none
fermented xylan.

Bacteroides isolates were separable into two
groups based on acidic fermentation products.
Most group 1 strains were aeroduric anaerobes
inasmuch as they could initiate growth in the
primary streak area of aerobic plates. However,
such growth was slow and sparse, and isolated
colonies were never formed. Group 1 isolates
also formed small amounts of propionate and
acetate from lactate during anaerobic growth in
PYL broth. However, growth in PYL was not

HE'I‘EROTROl’l-IIC BACTERIA IN TERMITE HINDGUTS

933

significantly greater than in medium lacking lac-
tate (i.e., PY broth). These observations sug-
gested that lactate fermentation by group 1
strains provides little or no energy for growth.
Nonetheless, to confimi the ability of such
strains to ferment lactate, experiments were per-
formed with cell suspensions. In fixed-time as-
says, two strains tested (strains S34C and
SSIOC) formed 0.52 to 0.70 iimol of propionate
and 0.29 to 0.3] pmol of acetate per umol of
lactate fermented per 2 mg of cell protein. No
lactate was fermented by heat-treated cell sus-
pensions.

Bacteria prescnt in guts of larvae and
soldiers. To determine whether bacteria pres-
ent in guts of workers were also present in other
colony members, isolatioiis were performed with
larvae and soldiers freshly collected from the
two sites. Results are presented in Table 3,
which includes data from workers for ease of
comparison. The number of viable bacteria per
gut was generally less than that for workers (see
also Table I), particularly in the case of first-
iiistar larvae, which were, in turn, considerably
smaller in size than workers. However, the major
groups of bacteria isolated from larvae and sol-
diers were strikingly similar to those obtained
from workers (see also Table 1). Streptococci
again represented the most abundant and con-
sistently isolatable group, and Bacteroides
strains were obtained in almost every instance.
All groups of isolates displayed characteristics
analogous to those described in Table 2.

Identification to species of streptococci
and Enterobacteriaceae. Because of the pre-
ponderance of streptococci among termite gut
isolates, an assessment of their phenotypic di-
versity seemed warranted. Accordingly, an at-
tempt was made to identify to species 145 strains
isolated from freshly collected termites (Table
3). Most streptococci (124/145) were identified

TABLE 2. General properties of bacteria isolated from guts of R. flavipes workers"

 

Gram reac-

Morphology “0"

Generic affiliation

 

Streptococcus C
Bacteroides
Group 1
Group 2
“Enteric type”
Lactobacillus
Fusobacterium
Unidentified

var.

2327232333;
|+l

+

Relation to

Acidii fermenta-

 

 

Catalase Motility . ‘
oxygen tion products
F(A) - - L(af)
AA(A) + - Al’siv
A + — apibbiv
F + + or — LAS
FtA) - - ' L(afs)

A ND — SABHP)
A ND — [apibbiv

 

“ Symbols: (+) Positive reaction; (—) negative reaction; (C) coccoid; (R) rod; (var.) variable; (F) facultative;
(A) strict anaerobe; (AA) aeroduric anaerobe; (ND) not determined. Acidic fermentation products: (A, a) acetic;
(L, l) lactic; (l’, p) propionic; (S, s) succinic; (B, b) butyric; (f) formic; (ib) isobutyric; (iv) isovaleric. Upper-case
and lower-case letters refer to major and minor amounts, respectively (19). Parentbeses indicate results with

some strains.
" See Table l, footnote d.

30

934 SCHULTZ AND BREZNAK

as either S. lactis or S. cremoris. The remainder
were represented by unidentified group N strep-
tococci and by red-pigmented strains whose
group affiliation was not determined. Curiously,
Janesville termites had a larger proportion of S.
lactis in their hindguts than S. cremoris,
whereas Spring Arbor termites had large num-
bers of S. cremoris but no demonstrable S. lac-
tis. Detailed physiological characteristics of the
streptococcal isolates are given in Table 4.
Attempts were also made to identify Entero-
bacteriaceae isolated from both Janesville and
Spring Arbor termites. Of 21 randomly selected

APPL. ENVIRON. MICROBIOL.

strains, 12 were indole-positive Citrobacter, 4
were citrate-negative Citrobacter, and 5 were
identified as Enterobacter cloacae. In addition,
none showed evidence of N2 fixation ability.

DISCUSSION

Results herein clearly show that an abundance
of viable, heterotrophic bacteria is present in the
hindgut of R. flavipes. Approximating the hind-
gut of workers to a right circular cone, with a
base radius of 0.4 mm and height of 4 mm (5),
its volume is about 0.7 mm“. Thus, the grand
mean number of viable cells per hindgut (3.5 x

TABLE 3. Bacteria isolated from guts of freshly collected larvae, soldiers, and workers of R. flavipes“

 

Generic affiliation of isolates

 

 

 

 

. :13): .No. f Streptococcus Bacteroides _
Bacteria collected “8 HI!" t isolates — “En- Uniden-
from: tharac- . Red- . Lacto- tified
(x ,, terized S we Unrden- pig- tent bactllus- anaero-
107 ) S. lactis ‘mu titied mented Group l Group 2 type'“ ' bic rods
group N strains
Janesville, Wis.
Larvae ’”
lst instar 0.4 25 14 1 3 1 2 O 0 4 0
3rd instar 1.6 23 11 3 0 2 0 2 2 2 1
Soldiers 2.0 30 9 4 3 0 4 4 2 2 2
Workers 2.1 44 18 6 4 0 14 I l 0 0
Spring Arbor,
Mich.
Larvae
lst instar 0.4 17 0 9 l 0 3 0 4 0 0
3rd instar 1.1 22 0 l3 4 0 4 0 1 0 0
Soldiers 1.0 19 0 1‘2 0 O 7 0 0 0 0
Workers 1.4 30 0 24 3 0 2 0 l 0 U

 

" Termites collected from Janesville and Spring Arbor in June and May. 1977, respectively. Isolation medium

was BHI in all cases.

" Average of two experiments (one using N 1/H, isolation atmosphere, the other using Gasl’ak) for each caste

or developmental stage.
" See Table 1. footnote d.

TABLE 4. Physiological reactions of Streptotoccus isolates"

 

 

Growth in milk

 

 

 

Growth in

l’recipitin reac-

 

 

 

Growth at: . ' BHI medium Hydrolysis of; . .
fl _ containing. . . tion With:
*— ‘ Heat __ _fl containing: Hemol ‘—_—‘ Ammo- ,,_
Isolate tolcr- "—_"" sis nia from
ance” 0.1%. 0.3".1 . , ,.. y. ' ., Hip- arginine Group Group
0 ‘ U ‘ 41),"! t).5 7' )N'U’ . ,
10 L 45 (J iiictliyl- iiietliyl- , , , . pur- N anti- 1) anti-
. Natl Natl lin
ene blue eno- lnue ate serum serum
S. cremoris + - — d — — — 7 — — _ + _
S. lactis + — — + + + -— y d — + + —
Unidenti- + - — + — + — a - — — + -
fied
group N
strains
Red-pig- — — - + — — — a — - +" N D N I)
mented
strains

 

“ Symbols: +, Positive; —, negative; d, different biochemical types (+, -, +“); +", weak positive; ND, not

determined.
" 60°C, 30 min.

31

VOL. 35, 1978

105) translates to 5.0 to 10" viable cells per ml of
hindgut fluid. This estimate is quite conserva-
tive, however, and probably low by 10- to 100-
fold, because the calculation ignores the appre-
ciable volume of the hindgut taken up by pro-
tozoa (5). Our estimate of 13% recovery of the
direct microscopic count is also conservative in
view of the method used to enumerate cells.
Homogenization of termite guts generates con-
siderable debris, derived not only from gut tissue
but also from disrupted protozoa. Yet during
microscopic counts, questionable particles were
scored as bacteria. Notwithstanding, there are
numerous bacteria with unique morphologies (5)
that were not isolated, e.g., spirochetes. Hope-
fully such organisms will be obtained in future
attempts, as our understanding of the hindgut
ecosystem increases and leads to medium for-
mulations that are more habitat-simulating.

The absence of cellulolytic bacteria in R. fla-
vipes guts is consistent with Hungate’s data (21 ),
which attribute cellulolysis in xylophagous ter-
mites to the protozoa. In addition, our inability
to demonstrate strict aerobes is consistent with
a concept of the hindgut being an anaerobic
niche (3). Although most of our isolates were
facultative anaerobes, they presumably use a
fermentative mode of energy generation in situ.
A challenge to this presumption is afforded by
Eutick et a1. (15), who concluded that the gut of
certain Australian termites was “aerobic.” While
their termite specimens were different from R.
flavipes, it should also be noted that these in-
vestigators did not measure the in situ p02, but
inferred the Eo’ of guts excised from termites fed
redox dyes. Consequently, extension of their
data to intact animals should be made with
caution.

Present findings parallel those of Krasil’nikov
and Satdykov (25), who isolated numerous
streptococci from Anacanthotermes species.
However, predominant streptococci in Anacan-
thotermes were strains of S. faecalis, whereas in
R. flavipes they were S. lactis and S. cremoris.
Our results present only an apparent contradic-
tion to those of Martin and Mundt (28). who
failed to demonstrate streptococci in R. flavipes.
The latter investigators used azide-containing
media to select for enterococci; azide inhibits the
growth of S. lactis and S. cremoris (24).

The preponderance of streptococci in R. fla-
vipes collected from widely separate sites, and
their persistence in the gut even after long cap-
tivity of the insects, suggests a stable relation-
ship between these bacteria and their host. How-
ever, any claim for autochthony (34) of strepto-
cocci in R. flavipes would be premature, since
some castes (winged forms and neotenics) re-
main to be examined. Presence of streptococci

HETEROTROPHIC BACTERIA IN TERMITE HINDGUTS

935

in guts of first-instar larvae, however, suggests
colonization early in the animals’ life. Acquisi-
tion of streptococci by larvae and soldiers is
probably via proctodeal food (31) solicited from
workers.

The abundance of streptococci in R. flavipes
hindguts, along with other bacteria that formed
lactic acid in pure culture (Table 2), suggests
that lactate could be a significant fermentation
product in situ and constitute an oxidizable en-
ergy source for the insect. Substrates for lactate
formation might be xylose and/or cellobiose,
which served as fermentable substrates for 145
streptococcal strains tested, and which could
arise during hemicellulose or cellulose hydrolysis
by other members of the biota (e.g., protozoa).
However, preliminary experiments (C. F. Hirsch
and J. A. Breznak, unpublished data) revealed
that no lactate was formed when suspensions of
hindgut microbiota were incubated anaerobi-
cally with [U-"C]cellulose. Relative amounts of
organic acids that were produced were: acetate
> formate > propionate > butyrate. Thus, either
lactate is not formed by streptococci in situ, or
it is formed but rapidly dissimilated by lactate
fermenters. The presence in guts of Bacteroides
strains (Tables 1 and 3), capable of fermenting
lactate to propionate and acetate, is consistent
with the latter suggestion and renders tenable
the hypothesis that interspecies lactate transfer
occurs between streptococci and Bacteroides.
This hypothesis is currently being examined in
our laboratory.

ACKNOWLEDGMENTS

We are grateful to Anna Findley, who participated in the
initial phase of this work, and to Duane Young for cooperation
during collection of termites in Spring Arbor, Mich. We also
thank Glen Esenther for helpful advice.

This investigation was supported by the Michigan Agricul-
tural Experiment Station of Michigan State University and
by National Science Foundation grants BMS75-05850 and
PCM77-05732.

LITERATURE CITED

1. Aranki. A., S. A. Syed. E. B. Kenney. and R. Freter.
1969. Isolation of anaerobic bacteria from human gin-
giva and mouse cecum by means of a simplified glove
box procedure. Appl. Microbiol. 17:568-576.

2. Beckwith, T. D., and E. J. Rose. 1929. Cellulose diges-
tion by organisms from the temiite gut. Proc. Soc. Exp.
Biol. Med. 27:4-6.

3. 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.

4. Breznak, J. A., W. J. Brill, J. W. Mertins, and H. C.
Coppel. 1973. Nitrogen fixation in termites. Nature
(London) 244:577-5tfl).

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

936

6.

12.

13.

18.

19.

Si

'21.

32

SCHULTZ AND BREZNAK

Bryant. M. P. 1974. Nutritional features and ecology of
predominant anaerobic bacteria of the intestinal tract.
Am. J. Clin. Nutr. 27:1313-1319.

. Cleveland, L. R. 1924. The physiological and symbiotic

relationships between the intestinal protozoa of ter-
mites and their host. with special reference to Reticu-
litermes flavipes Kollar. Biol. Bull. 48:178—227.

. Conn, H. J., M. W. Jenniaon, and 0. B. Weeks. 1957.

Routine tests for the identification of bacteria. p.
140-168. In H. J. Conn and M. W. Jennison (ed.),
Manual of microbiological methods. McGraw-Hill Book
Co.. Inc., New York.

. Deibel, R. H. 1964. Utilization of arginine as an energy

source for the growth of Streptococcus faecalis. J. Bac-
teriol. 87:988-992.

. Deibel, R. H., and H. W. Seeley, Jr. 1974. Genus I.

Streptococcus Rosenbach 1884, 22. In R. E. Buchanan
and N. E. Gibbons (ed.), Bergey’s manual of determi-
native bacteriology, 8th ed. The Williams 8:. Wilkins
00., Baltimore.

. Dickman. A. 1931. Studies on the intestinal flora of

termites with reference to their ability to digest cellu-
lose. Biol. Bull. 61:85-92.

Dworkin. M., and J. W. Foster. 1958. Experiments with
some microorganisms which utilize ethane and hydro-
gen. J. Bacteriol. 75:592-603.

Edwards, P. R., and W. H. Ewing. 1972. Identification
of Enterobacteriaceae, 3rd ed. Burgess Publishing Co..
Minneapolis.

. Esenther. G. R. 1969. Termites in Wisconsin. Ann. En-

tomol. Soc. Am. 62:1274-1284.

. Eutick, M. L., R. W. O’Brien, and M. Slaytor. 1976.

Aerobic state of gut of Nasutitermes exztiosus and
('optotermes lacteus. high and low caste termites. J.
Insect Physiol. 22: 1377-1380.

. Francisco, D. E., R. A. Mah, and A. C. Rabin. 197:1.

Acridine orange-epifluorescence technique for counting
bacteria in natural waters. Trans. Am. Microsc. Soc.
92:416-421.

. French, J. R. J. 1975. The role of termite hindgut bacteria

in wood decomposition. Mater. Org. 10:1-13.

French, J. R. J., G. L. Turner, and J. F. Bradbury.
1976. Nitrogen fixation by bacteria from the hindgut of
termites. J. Gen. Microbiol. 95:202-206.

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

. Hungate. R. E. 1936. Studies on the nutrition of Zooter-

mopsis. I. The role of bacteria and molds in cellulose
decomposition. Zentralbl. Bakteriol. Parasitenkd. Infek-
tionskr. Abt. 2 94:240-249.

Hungate. R. E. 1939. Experiments on the nutrition of
Zootermopsis. Ill. The anaerobic carbohydrate dis-
similation by the intestinal protozoa. Ecology
20:230-245.

. Hungate, R. E. 1946. Studies on cellulose fermentation.

II. An anaerobic cellulose-decomposing actinomycete,

24.

27.

31.

3‘2.

Ilt'.

U

37.

APPL. ENVlItON. MICROBIOL.

Mtcromonospora pr0pionici, n. sp. J. Bacteriol.
51:51-56.

. 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.

Kenner, B. A., H. F. Clark. and P. W. Kabler. 1961.
Fecal streptococci. I. Cultivation and enumeration of
streptococci in surface waters. Appl. Microbiol. 9:15-20.

. Krasil'nikav, N. A., and S. I. Satdykov. 1970. Bacteria

of termites’ intestines. Microbiology 39:562-564.

. brunette. E. H., E. H. Spaulding, and J. P. 'I‘ruant

(ed.). 1974. Manual of clinical microbiology. 2nd ed.
American Society for Microbiology, Washington, DC.

Mannesmann. R. 1972. A comparison between cellulo-
lytic bacteria of the termites Coptotermes formosanus
Shiraki and Reticulitermes virgtntcus (Banks). Int.
Biodeterior. Bull. 8:104-111.

. Martin. J. D., and J. 0. Mundt. 1972. Enterococci in

insects. Appl. Microbiol. 24:575-580.

. Mauldin. J. R. 1977. Cellulose catabolism and lipid syn-

thesis by normally and abnormally faunated termites,
Reticulitermes flavipes. Insect Biochem. 7:27-31.

. Misra. J. N., and V. Ranganathan. 1954. Digestion of

cellulose by the mound building termite. Termes (Cy-
clatermes) obesus (Rambur). Proc. Indian Acad. Sci.
39:100-113.

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.

Pochon. J., H. de Barjac, and A. Roche. 1959 Re-
cherches sur la digestion de la cellulose chez le termite
Sphaerotermes sphaerothorax. Ann. Inst. Pasteur Paris
98:352-355.

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

fixing Enterobacter agglomcmns isolated from guts of
wood-eating termites. Appl. Environ. Microbiol.
33:392-399.

. Savage. D. C. 1977. Microbial ecology of the gastrointes-

tinal tract. Annu. Rev. Microbiol. 31:107-133.

. Sebald. M., and A. R. Prévot. 1962. Etude d'une nou-

velle espec anaerobie stricte Mtcromonospora aceIo/or-
mu-i n. sp. isolee de l'intestin posterieur de RPltt'uIl-
termes Iuctfugus var. samtonnensts. Ann. Inst. Pasteur
Paris [02:199-214.

Speck. U., G. Becker. and M. Lena. 1971. Ernahrungs-
physiologische Untersuchungen an 'l'ermiten nach se.
lektiver medikamentoser Ausschaltung der Darmsym-
bionten. Z. Angew. Zool. 58:475-491.

Tetrault, I’. A., and W. L. Weis. 19:17. Cellulose (ll-1‘01":
position by a bacterial culture from the intestinal tract
of temiites. J. Bacteriol. 33:95.

. Thayer, D. W. 1976. Facultative wood-digesting bacteria

from the hind-gut of the termite Reticulitermes hes-pc-
rus. J. Gen. Microbiol. 95:287-296.

ARTICLE 2

CROSS-FEEDING OF LACTATE BETWEEN STREPTOCOCCUS LACTIS

 

AND BACTEROIDES SP. ISOLATED FROM TERMITE HINDGUTS

 

By. J. E. Schultz and John A. Breznak

33

ABSTRACT

Streptococcus lactis and Bacteroides sp., isolated from hind-

 

 

guts of Reticulitermes flavipes termites were grown anaerobically in

 

monoculture and coculture. When grown in a glucose medium, §, lag§i§_
monoculture produced lactate as the major fermentation product with
small amounts of formate, acetate, ethanol, and C02. In coculture,
glucose was completely consumed during growth of §, lactis, Lactate,

produced by §, lactis, then supported growth of Bacteroides and was

 

fermented to pr0pionate, acetate, and C02. Small amounts of succin-

ate were formed during growth of Bacteroides in the coculture, but

 

little change in the formate or ethanol concentration was observed.

Monoculture growth of Bacteroides in a tryptone-yeast extract medium

 

revealed that small amounts of lactate (20 to 40 nmoles/ml) increased
cell yields and production of ogranic acids. However, initial lactate
concentrations greater than 40 nmoles/ml suppressed not only growth

of Bacteroides but acidic product fOrmation as well. Results suggest

 

that cross-feeding of lactate between streptococci and bacteroides
constitutes one aspect of the overall hindgut fermentation in the

termite.

35

INTRODUCTION

The hindgut of wood-eating termites contains an abundant,
heterogeneous population of bacteria whose roles in carbon, nitrogen,
and energy metabolism are not well-understood. Isolation and charac-

terization of heterotrOphic bacteria from hindguts of Reticulitermes

 

flavipes (family Rhinotermitidae) has revealed a predominance of

streptococci and Bacteroides species which appear to be relatively

 

stable components of the hindgut microbiota (14). Almost all of the
streptococci isolated from B, flavipes were §, lactis and §, cremoris.
They were saccharolytic and displayed a homolactic fermentation of

glucose. Bacteroides isolates were also saccharolytic, but selected

 

strains additionally possessed the ability to ferment lactate to pro-
pionate and acetate as major soluble products (14).
These findings suggested that cross-feeding of lactate from

streptococci to Bacteroides might be one feature of the overall hind-

 

gut fermentation occurring in the insect. To test the feasibility of

this hypothesis, strains of s, lactis and Bacteroides sp. were ex-

 

amined during in_vitro culture. In this paper we report on the
growth, substrate utilization, and product formation by §, lactis

in monoculture and in coculture with Bacteroides sp. In addition,

 

the effect of lactate on growth and product formation by pure cul-

tures of Bacteroides was examined.

 

36

MATERIALS AND METHODS

Organisms. Streptococcus lactis strain 0H] and Bacteroides

 

sp. strain JH20 were isolated from hindguts of Reticulitermes flavipes

 

workers collected from Janesville, Hi. (14). Strain szo was assigned

to group 1 on the basis of acidic fermentation products (l4).

Media and growth conditions. Strict anaerobic techniques (5)

 

were used for preparation of media, growth of cells, and sampling of
culture contents. Growth media were TY, TYC, and TYG. TY medium con-

tained (g/l00 ml): tryptone, l.0; yeast extract, 0.l; (NH4)2504, 0.2;

4 4

KH P0 ; and resazurin, l0' .

2 4’
The initial pH was 7.0. TYC was identical to TY, but also contained

0.34; NaZHP04, 0.66; hemin, 5 x lo.

cysteine-HCl (0.05 g/l00 ml). TYG was identical to TY, but also con-

140 was used to TYG

tained glucose (0.1 g/l00 ml). When glucose-UL-
medium, it was sterilized separately by filtration and incorporated
at an activity of 0.5 uCi/ml.
Cells were usually grown in 300 ml Nephelo flasks (Bellco

Glass Inc., Vineland, N.J.) equipped with a serum-stoppered sampling
port and containing 100 to 200 ml of medium. However, for some ex-
periments cells were grown in rubber-stoppered l8 x l42 mm anaerobe
tubes (Bellco) containing l0 ml of medium. The initial gas phase in
all cultures was 02-free N2. Inocula consisted of exponential phase

cells growing in homologous media, and the incubation temperature

was 20°C.

Analysis of growth and fermentation products. Growth of cells

 

was monitored turbidimetrically by measuring the absorbance of cultures

37

at 660 nm with a Spectronic 20 colorimeter (Bausch & Lomb). In addi-
tion, viable cell counts were determined by plating appropriate dilu-
tions of culture onto plates of supplemented brain heart infusion
(BHI) agar and incubating them in GasPak jars as previously described
(14). Under these conditions, colonies of §,'uyggE;le were 5 mm
in diameter and round with a slightly erose margin, whereas those of

Bacteroides JWZO were only l.5 mm in diameter and were round with an

 

entire margin. Thus, viable cells of each strain could be easily
quantified, even when growing together in the same vessel (i.e. in
coculture; Fig. l).

For time-course analyses of fermentations, culture samples were
centrifuged at 16,300 x g for l0 min at 10°C, and the resulting super-
natant fluids were frozen until ready for assay. Glucose and lactic
acid were assayed by the methods of Spiro (l5) and Barker (2), re-
spectively. The molecular configuration of lactic acid was determined
enzymatically (5). Other organic acids were quantified by using gas
chromatography (l3). Ethanol was estimated enzymatically (Ethyl
Alcohol Reagent Set; Worthington Biochemical Corp., Freehold, N.J.).
Results with TYG medium were corrected for small amounts of products
produced during growth of cells in the absence of glucose (i.e. in
TY medium).

14C were determined after

Products formed from glucose-UL-
cells had grown for 8l h. At that time a small volume of culture was
removed for determination of viable cell number, and the remainder
of the culture was acidified with H2504 to a final concentration of
0.12 H, The headspace gas was then swept into a phenethylamine/

methanol (lzl, vol/vol) trap by using N2 as carrier. Radioactivity

38

Figure l. Surface colonies of Streptococcus lactis strain JNl
(large colonies) and Bacteroides sp. strain JN20
(smaller colonies) on a l00 mm diameter plate of
supplemented BHI medium. Incubation was at 25°C for
5 days.

 

39

 

Figure l.

4O

14

trapped in this way was assumed to be due solely to C02, whose

14C.

quantity was calculated from the specific activity of glucose-UL-
The acidified fermentation liquor was then clarified (ll) and assayed
as described above. In addition, organic acids were extracted from the
4 clarified liquor with ether and were separated by silicic acid column
chromatography (ll) for determination of specific radioactivity. For
determination of 14C incorporation into cell material, cells were
harvested by centrifugation, hydrolyzed with NaOH (4), then neutralized
with HCl and l M phosphate buffer (pH 7.0) before determination of
radioactivity.

Radioactivity of samples was determined by using a Delta model
300 liquid scintillation spectrometer- (Searle Analytic Inc., Des
Plaines, Ill.) and ACS scintillant (Amersham, Arlington Hts., 11].).
All samples were corrected for self absorption of radiation by the

channels ratio method (17); toluene-14C was used as a radioactivity

standard.

Chemicals. Tryptone and yeast extract were obtained from

14 14c

Difco Laboratories, Detroit, Mi.; glucose-UL- C and toluene-
from New England Nuclear, Boston, Ma.; and sodium-DL-lactate (syrup)
from Sigma Chemical Co., St. Louis, Mo. All other chemicals were of

reagent grade and were obtained from commercial sources.

41

RESULTS

Growth of S. lactis JWl and Bacteroides sp. JW20 in monoculture

 

and coculture. S, lactis grew readily as a monoculture in TYG medium,
l

 

exhibiting specific growth rates of about 2.0 x hr-

ing from 2.0 x 108 to 2.5 x 108 viable cells per ml. During growth

and yields rang-

glucose was consumed and lactic acid production ranged from 7.6 to

8.0 umole/ml, thereby accounting for 69 to 72% of the glucose carbon
fermented. The molecular configuration of lactic acid was L(+). These
events were completed during the first 10 to 15 h of incubation; there-
after the number of viable cells and lactic acid concentration remained
constant for up to 80 h. Small amounts of ethanol, acetate, formate,
and C02 were also produced, and about 6% of the glucose carbon was
incorporated into cell material (Table 1; column 1). When glucose-
UL-14C was used as substrate, the specific activity of lactate pro-
duced indicated that virtually all of the acid was derived from glu-

cose (Table l; footnote a).

When S, lactis JWl was cocultured with Bacteroides JW20, the

 

first 10 h of incubation was dominated by growth of JWl and by glucose
consumption and lactate production (Fig. 2). Thereafter, strain JW20
commenced growth at a specific rate of about 0.5 x hr"1 and achieved

8 to 8.6 x 108 viable cells per m1.

populations ranging from 6.0 x 10
During the latter stages of growth in particular, lactate previously
produced from strain JWl decreased markedly while propionate and
acetate concentration increased (Fig. 2; acetate not depicted). C02
and small amounts of succinate were also formed during growth of

Bacteroides in the coculture, whereas little change in formate or

 

42

Table 1. Fermentation of glucose-UL-14C by S, lactis JWl in mono-
culture and in coculture with Bacteroides sp. JW20.

 

 

 

 

 

 

 

 

 

umol /100 umol gl ucose-UL-MC ulnaoclt/ilcooalcfli'g]
fermented fermented
Productsa
S, lactis + Coculture Coculture
S, lactis Bacteroides minus minus
monoculture coculture monoculture monoculture
Lactate 138.5 36.0 -102.5 --
Propionate 0.0 80.9 80.9 78.9
Acetate 7.2 26.9 19.7 19.2
C02 3.6 30.6 27.0 26.3
Ethanol 12.6 9.8 -- --
Formate 5.3 5.4 -- --
Succinate 0.0 2.7 -- --

Carbon recovery (%)

in prOdUCtS 77.3 78.4 100.5 100.5
in censb 5.8 6.1 -- __
Total carbon 83.1 84.5 -- --

recovery (%)

 

aSpecific activity of glucose-UL-14C substrate was 31,500
dpm/umole carbon. Specific activities of lactate in the monoculture
and coculture were 30,550 and 28,000 dpm/umole carbon, respectively.
Specific activities of propionate and acetate in coculture were
22,540 and 20,260 dpm/umole carbon, respectively.

bDetermined in separate experiments under identical condi-
tions.

43

Figure 2. Growth, substrate utilization, and product formation by a
coculture of Streptococcus lactis strain JWl and Bacteroides

 

 

sp. strain JW20. The growth medium was TYG.

44

(O) |w/ sqoum

 

 

.N mg=m_d

3:0:

 

 

(0) IW/Suoa qan 601

45

ethanol concentration was observed (Table 1; column 2). When the
quantity of products formed by S, lggti§_monoculture was subtracted
from that produced in coculture, then normalized to lactate consump-
tion, propionate, acetate, and C02 fbrmation by strain JW20 accounted
fOr 100.5% of the lactate carbon dissimilated (Table 1; columns 3 and
4). The specific activities of propionate and acetate in the coculture
were 81 and 72%, respectively, that of lactate carbon (Table 1; foot-
note a). These data indicated that lactate was the major source of
propionate and acetate during growth of JW20, although a slight dilu-
tion of label occurred. Nevertheless, the apparent stoichiometry of
lactate dissimilation by strain JW20 (100 lactate--+ 79 propionate +
l9 acetate + 26 C02) agreed reasonably well with that anticipated

(100 lactate--+-66 propionate + 33 acetate + 33C02) fbr the dissimila-
tion of lactate by either the acrylate (16) or succinate pathway (9).

Bacteroides JW20, grown as a monoculture in TY medium (i.e.

 

lacking added glucose or lactate), exhibited a growth rate of 0.3 x

hr-1

and yielded only 2.0 x 108 viable cells per ml. Likewise, only
small amounts of acetate (0.20-0.23 nmoles/ml) and propionate (0.50-
0.60 nmoles/ml) were formed. The addition of cysteine to TY medium
(i.e. TYC medium) had no effect on the cell yield, but resulted in a

faster growth rate comparable to that of Bacteroides in the coculture.

 

Comparison of cell yields indicated that growth of JW20 was 4-fbld
greater in cocultures that in TY or TYC monocultures. From these data
we conclude that lactate, produced by S, lactis JWl, was used as an

energy source for growth of Bacteroides JW20.

 

46

Effect of lactate ongrowth of Bacteroides JW20 in monoculture.
The importance of lactate as an energy source for strain JW20 in co-
culture prompted an examination of its effect on JW20 in monoculture.
Results are presented in Table 2. Incorporation of small amounts of
lactate (20 to 40 nmoles/ml) in TYC medium increased cell yields 3-fold
and markedly increased production of propionate, acetate, and succin-
ate. However, higher initial levels of lactate supressed not only
final cells yields, but organic acid production as well. At an initial
concentration of 120 nmoles/ml, the cell yield was scarcely above that
obtained with TYC medium alone. By contrast, incorporation of 11 mM
glucose into TYC medium resulted in cell yields greater than those
achieved with lactate and also shifted the fermentation pattern in
favor of succinate production. The apparent growth inhibition by
lactate at concentrations 3_60 mM was not due to a loss of the buffer-
ing capacity of the medium; final pH values of all cultures ranged from

6.5 to 6.9 (Table 2; footnote a).

Table 2. Effect of lactate on growth and product farmation by
Bacteroides JW20 in monoculture

 

47

 

Maximum Yielda

Products (nmoles/m1)

 

 

 

 

 

 

 

Addition to

TYC medium

(nmoles/m1) 0.0.560 :2:D;$ fi1138) Propionate Acetate Succinate

No addition 0.12 2.0 0.60 0.20 0.10

Sodium-OL-

lactateb
20 0.31 5.7 11.43 2.89 0.49
40 0.33 6.1 15.85 3.91 0.50
60 0.26 4.8 12.00 3.14 0.38
80 0.21 3.8 9.14 2.41 0.33
100 0.18 3.2 7.51 1.97 0.25
120 0.14 2.4 5.81 1.60 0.13

Glucose, 11 0.82 15.5 1.51 1.18 4.10

 

aA11 values far optical density (0.0.) were maximum at termin-
ation (120 h), except for TYC with glucose and TYC alone which at-
Final pH values

tained maximum 0.0.550 at 24 and 58 h, respectively.
ranged from 6.5 to 6.9.

bThe molarity of commercial sodium-DL-lactate used to prepare
the media was determined by assay (2) against standard zinc lactate.
The amount of lactate remaining at termination was not determined.

 

1m11111111111111»11111111111111 £5

49019