Antibiotics

That Come The differences in the types of plants

investigated, the available quantities
of particular varieties, and the kinds
From Plants of equipment used for processing mean
that the methods of extracting plant
antibiotics are almost as numerous as
P. S. Schaßer, William E. Scotts the groups of scientists here and abroad
Thomas D. Fontaine who are investigating them. The or-
ganisms against which the extracts are
tested are varied and many, too.
As to procedure, the investigators
An antibiotic is an organic chemical disagree. One group believes that only
substance that is produced by a plant, fresh green plants should be used, be-
an animal, or a micro-organism. It cause, they say, plants should be har-
selectively checks the growth of bac- vested quickly and handled immedi-
teria, viruses, fungi, and other disease- ately before enzymatic processes can
producing organisms, or completely de- destroy the antibiotic substances or pro-
stroys them. The term is new, but the duce toxic byproducts. In theory, that
idea that such substances can cure dis- method is preferred; it can be con-
eases is as old as the hills. ducted successfully on a small scale.
Investigations since 1943 have dis- Another group works almost entirely
closed that many plants, plant parts, with dried material; the commercial
and plant extracts have therapeutic preparation of an antibiotic would al-
values (a point that primitive man most always require the drying or
knew) and contain antibiotic agents longer handling of the plant material.
(a point that primitive man might have In our laboratory, we use some fea-
suspected). Among them are the lotus, tures of both methods. We first assay
olive, laurel, myrtle, asphodel, and gar- the extracts of freshly cut green plants
lic, which were used as medicines in and, having detected antibiotic activ-
ancient Assyria ; dates, figs, onions, let- ity, we then dry another portion in a
tuce, crocus, and opium, used as rem- forced-draft oven, usually at 85° C,
edies by the Egyptians; the juices of for extraction. If the antibiotic ma-
celery, parsley, asparagus, peppers, and terial is not destroyed by the drying
cabbage, favored as a medicine by process, we dry a quantity sufficient for
Greeks and Romans ; and amber, musk, work during the winter, when the fresh
manna, cloves, peppers, rhubarb, nut- plants may not be available. We have
meg, camphor, crotón oil, and nux found it desirable, if possible, to sepa-
vómica, which the Arabians intro- rate the plants into their various
duced and told their neighbors to take parts—roots, leaves, stems, and flowers
for aches and pains. —since the antibiotic may be present
Only a few antibiotics so far have in one or more but not in all parts of
been isolated from the higher plants in the plant. Plants may be extracted with
pure form. The detection, isolation, water, saline solution, dilute acid or
characterization, and evaluation of alkali, or organic solvents, such as alco-
antibiotics require the coordinated ef- hol, acetone, and ether. The use of
forts and experiences of plant physiol- lower-boiling solvents is preferred in
ogists, bacteriologists, chemists, and order to avoid excessive heating when
medical men. concentrating the extracts.
727
728 1950-19 o 1 YEARBOOK OF AGRICULTURE

The detection and measurement of (Escherichia) coli as test organisms,

antibiotic activity can be accomplished and found that 63 genera contained
by a number of methods. substances that inhibited the growth of
In the cylinder-cuj-i and paper-disc one or both test organisms.
methods, a Pctri dish containing a solid In the botanical and bacteriological
nutrient agar is inoculated evenly with laboratories at Indiana University,
the micro-organism. In the first, cylin- D. W. Sanders and his associates tested
ders, made of porcelain, glass, or stain- the juices from 120 plant species, col-
less steel, are then placed on the inocu- lected mostly in Indiana, using Es-
lated agar, and the plant extract is cherichia coli and Bacillus subtilis as
pipetted into the cylinders. In the other test organisms. Twenty-two of the
method, paper discs are immersed in plants showed varying degrees of bac-
the plant extract long enough to be- terial inhibition.
come saturated with the solution, after H. J. Carlson and his associates at
which they are placed on the inocu- Western Reserve University extracted
lated agar. The agar plate is incubated and tested more than 200 plants, col-
for a fixed period of time at a tem])era- lected in Oregon, in an attempt to
ture that is optimum for the growth of separate substances inhibitory to ma-
the organism. Any zones of inhibition laria parasites, bacteria, and viruses.
around the cylinders or discs are meas- Alany of them were found to contain
ured to obtain a relative indication of substances that were bacteriostatic or
the amount of antibiotic agent in the bactericidal to micro-organisms in
extract. vitro, i. e., in laboratory test. Extracts
The serial dilution and streak (or prepared from five of the plants—
transplant) methods diiïer from the buttercup, sagcbi'ush, the mountain
first two mainly in that graded pasque, tiwarf waterleaf, and juniper—
amounts of the plant extract to be were evaluated by many comprehen-
tested are incorporated in the nutrient sive tests in vivo (i. c., in experimental
medium, and the micro-organism is animals) and by in vitro tests. Salt ex-
then added. tracts of all five plants wx^re found to
We use all of these methods in our have antibacterial and antimalarial
survey work and in purification of anti- activity in vitro. Two of the plants,
biotic agents present in plant extracts. sagebrush and dwarf waterleaf, con-
We find it advantageous to use rou- tained substances that protected chick-
tinely several typical test organisms— ens during the blood phase of malaria,
the gram-positive bacterium Staphy- and an ether-insoluble, water-soluble
¡0coccus aureus, the gram-negative fraction of a steam distillate of moun-
bacterium Escherichia coli, the fungus tain pasque protected mice heavily in-
Fiisarium oxyspomm f. lycopersici, and fected with pncumococcus organisms
the acid-fast bacterium Mycohacte- {Diplococcus pneumoniae type 19).
rium phlei. Plant extracts may inhibit Extracts obtained from a species of
the growth of one or more or all of the sumac, Rhus hirta, showed marked
organisms, but usually when a crude bacteriostatic activity against gram-
plant extract inhibits the growth of all negative bacteria but were less effec-
four the presence of more than one tive in inhibiting the growth of gram-
antibiotic in the plant is indicated. positive bacteria and fungi.
The widespread antibiotic activity L. E. Hayes studied 231 plant spe-
in plants is attested by the published cies, collected in Ohio, as possible
reports of various investigators. At Ox- sources of antibiotics. Extracts from 46
ford University, E. M. Osborn in 1943 inhibited the growth of one or more
tested the antibiotic activity of water species of the test organisms—Staphy-
extracts from 2.300 species of plants iococcus aureus, Escherichia coli, Er-
that belong to 166 families; she used winia carotovora, and Phytomonas
Staphyiococcus aureus and Bacterium turne ja ci ens.
ANTIBIOTICS THAT COME FROM PLANTS 729
About 100 plants listed in the Indian tracts, and of these approximately 50
Pharmacopoeia for the treatment of percent showed inhibition against one
diseases that are definitely of bacterial or more of the representative gram-
origin have been found by Mariam positive, gram-negative, and acid-fast
George and associates of the Indian bacteria or fungi.
Institute of Science to contain anti- In the preliminary survey of plants
biotics. R. R. Rao and associates re- for antibiotic activity, we use methanol
ported that garlic extract in low con- as a solvent in the preparation of ex-
centration was bacteriostatic in vitro tracts. The methanol extract is evapo-
against the Mycobacterium tuberculo- rated to dryness on a water bath under
sis organism. reduced pressure, and water is added
At the University of Vermont, to the residue until 1 milliliter of water
Thomas Sproston, Jr., and associates extract represents 1 gram of dry plant
tested 73 plant extracts, directing par- material. Any water-insoluble material
ticular attention toward those that is removed by centrifuging, and the
show fungicidal or fungistatic proper- supernatant solution is poured off and
ties. The most active antifungal ex- tested for antibiotic activity. If any is
tracts were obtained from Impatiens found, the stability of the antibiotic
(wild touch-me-not), Cucumis nielo L. principle or principles to heat is de-
(muskmelon), and Tropaeolum majus termined by autoclaving a portion of
L. (nasturtium). The crystalline anti- the aqueous extract for 15 minutes at
fungal agent isolated from Impatiens 15 pounds steam pressure.
was 2-methoxy- 1,4-naphthoquinone.
At Columbia University, B. C. THE MANY UNIDENTIFIED antibiotic
Seegal and M. Holden found that the principles of the higher plants vary
pressed juice or the steam distillate greatly, not only in their potency and
from the pressed juice of buttercup is distribution within the plant but also
a strong antibiotic with a wide range M.dth the species and variety of plant.
of activity. The extracts were effective For instance, E. H. Lucas and R. W.
in vitro against selected gram-positive Lewis at Michigan State College re-
and gram-negative pathogenic cocci ported that the active principles found
and bacilli, against Mycobacterium in the scarlet berries of Lojiicera
tuberculosis, and against three yeasts, tatarica, one of the honeysuckles, were
two of which are potential human not present in species and varieties of
pathogens. The toxicity of the active honeysuckles having dark-red, orange,
principle, protoanemonin, however, is yellow, or purple fruits. G. S. Pederson
sufficient to preclude its use as an and Paul Fisher, in their investigation
effective therapeutic agent. of the bactericidal action of cabbage
At the New York State Agricultural juice, found that four of the varieties
Experiment Station, C. S. Pederson tested showed very marked bactericidal
and Paul Fisher noted that the unde- action, five showed intermediate effect,
sirable gram-negative aerobic bacteria and four showed still less effect. E. M.
on the surface of cabbage leaves ordi- Osborn concluded from her survey of
narily disappear shortly after the 2,300 plants that specificity and activ-
cabbage is cut. They attributed this ity of plant antibiotics were similar
change in the number of micro-organ- among members of the same family.
isms to the presence of bactericidal Our own tests on unpurified plant ex-
substances in the cabbage tissue, which tracts did not indicate any great dif-
caused marked reduction in the nuni- ference among species and varieties of
ber of gram-negative bacteria in 6 to plants in the same family. There may
24 hours. The extract was less active be differences in antibiotic activity due
toward gram-positive bacteria and was to climatic conditions, but, so far as wc
inactivated by heat. know, none of the investigators in the
Wc have tested some 300 plant ex- field has studied this phase intensively.
 50-1951 YEARBOOK OF AGRICULTURE
730
The ultimate goal of any scientist Ci tHinOj, and it is designated as a beta,
working on the production of antibi- gamma unsaturated lactone. The crys-
otic substances from plant materials is tals are polymorphic, soluble in alco-
the preparation of the active principle hol and ether, very slightly soluble in
in a pure and crystalline form. The water, heat stable, acid stable, and al-
structure and biological properties of kali unstable. Crepin is most active
the crystalline antibiotic can then be against gram-positive bacteria, but its
determined, and its synthesis in the lab- activity is diminished by scrum. It de-
oratory may possibly be accomplished. stroys the motility of leucocytes at a
The method of isolation varies with concentration of 1 :450,000 and is le-
the nature of the antibiotic. Some of thal at higher concentrations.
the substances are acidic, some neutral, Another antibiotic ]3rinciple was iso-
some basic, and some proteinaceous. lated from Spiraea aruncus L. (goat's
The compounds in the acidic group beard). This crystalline compound is
can be extracted at low pH by organic thought to be an alpha, beta unsatu-
solvents immiscible w^th water. The rated lactone with a suggested formula
basic compounds can be precipitated CjoHiiOf. It has very low activity
by base précipitants. Some antibiotics against gram-]Dositivc and gram-nega-
are purified by extracting the plant tive bacteria.
material with one solvent and then dis- C. J. Cavallito and J. H. Bailey of
tributing the crude extract between the Sterling-Winthrop Research Insti-
two immiscible solvents. The plant tute have isolated antibiotic agents
material may also be extracted with a from the following plants: Arctium
suitable solvent, after which the extract mÁnus (common burdock). Asarum
is concentrated and run through a canadense (wild ginger), Alliuin sati-
Chromatographie column containing vu7n. (garlic), and Centaurea macu-
an adsorbent capable of removing the losa (spotted knaj^weed). A crystalline
active substance. The antibiotic may antibiotic from Arctium minus was
then be removed from the adsorbent obtained by fractionation of a water
wâth a suitable solvent, purified, and extract and subsequent crystallization
crystallized. The pure crystalline com- from ethyl acetate. The suggested for-
pound may be identified by determin- mula for this compound is Ci.-^HjoO^,
ing its chemical constituents and and it probably is an unsaturated lac-
ascertaining its physical and chemical tone. The crystals are colorless prisms
properties. Determination of the spec- soluble in alcohols, chloroform, ethyl
trum, color reactions, with various re- acetate, dioxane, and acetone; slightly
agents, sulfhydryl inactivation, and soluble in water; insoluble in j^etroleum
other special methods are useful in ether; optically active; melting at
characterizing an antibiotic. 115° to 117° G. and resolidifying to
lose antibiotic activity. The antibiotic
THE FOREGOING description of the is irreversibly inactivated by cysteine.
procedure for isolating a pure antibi- It is active only against gram-positive
otic has necessarily been very general, bacteria. The LD,-,i—that is, the
as the methods vary with the com- amount that will kill 50 percent of the
pound isolated. A few examples will be animals—for mice is 90 milligrams
cited briefly. per kilogram intravenously. The sub-
In England, N. G. Heatley and as- stance is not lethal but is toxic at 500
sociates isolated a crystalline antibiotic mg./kg. when administered orally.
from Crépis taraxacifolía by adsorp- Two antibiotic principles were iso-
tion of the active principle from an lated from Asarum canadense (wild
aqueous extract on charcoal and fur- ginger). They are designated as com-
ther purified it by adsorption on alu- pounds A and B, with the suggested for-
mina. The suggested chemical formula mulas Cr_>H,oOsN,S and Gir.H„0:N,
for the antibiotic, named crepin, is respectively. Both compounds are al-
ANTIBIOTICS THAT COME FROM PLANTS

most insoluble in water, benzene, and among these are three plant-wilt or-
petroleum ether, but are soluble in al- ganisms—Fusarium oxysporum f. lyco-
cohol, acetone, chloroform, ethyl ace- persici (tomato wilt), Fusarium oxy-
tate, and dioxane. Both are active only sporum i. pisi (pea wilt), and Fusarium
against gram-positive bacteria. Prelim- oxysporum f. conglutiiians (cabbage
inary toxicity tests indicate A to be yellows)—and six fungi that are path-
lethal in about 3 days after intraperi- ogenic to human beings—Blastomyces
toneal injection of 5. mg./kg. in sesame dermatitidis.Coccidioides immitis, His-
oil. toplasma capsulatum, Epidermophyton
Allium sativum (garlic) yielded a floccosum, Trichophytou mentagro-
sulfur-containing antibiotic, which was phytes, and Candida albicans.
equally active toward gram-positive Fractionation of the tomatin extract
and gram-negative bacteria. The com- has resulted in the separation of an
pound w^as isolated as a colorless liquid, individual crystalline antibiotic, to-
stable to acid but unstable to heat and matine, which has antifungal activity
alkali, soluble in petroleum ether, and but little or no antibacterial activity.
irreversibly inactivated by cysteine. Riitin, a flavone glycoside, also has
The exact formula for this compound been isolated from the extract. Rutin
is not known, but two are suggested: docs not have antibiotic activity, but
GH.=CH - CH. - S - S - CH2 - GH=CH2 its degradation product, quercetin,
II does have antibacterial properties.
O Therefore^ we assume that at least part
or of the antibacterial activity exhibited
GH.=GH-CH2-S-0-S-GH2-CH=CH.
by tomatin is due to the presence of a
In toxicity tests with mice, the LD50 small amount of quercetin.
was 60 mg./kg. intravenously and 120 Tomatine was isolated from a crude
mg./kg. subcutaneously. tomatin concentrate by repeated pre-
The dried leaves of the spotted knap- cipitation from an alkaline solution,
weed^ Centaurea maculosa, yielded dissolving the alkaline precipitate in
about 1.5 percent of an antibiotic hot 70-percent ethanol, and cooling,
which is active against gram-positive thereby precipitating an amorphous
and gram-negative bacteria. The sug- substance, which was dried and then
gested formula for this unsaturated dissolved in hot 80-percent dioxane.
lactonc is GooH^fjOj. Its activity is rap- Crystalline tomatine w^as obtained on
idly destroyed by cysteine and thio- cooling the dioxane solution.
glycolate. Tomatine has been characterized as
Two compounds, lupulon and hum- a glycosidal alkaloid. It is soluble in
ulon, isolated from the resins of mature ethanol, methanol, dioxane, and pro-
hops, were recently found to have anti- plene glycol; it is almost insoluble in
biotic activity toward gram-positive w^ater, in ether, and in petroleum ether.
and acid-fast bacteria {Mycobacterium It appears to be stable in strong alkali,
tuberculosis) in research on hops utili- but is readily hydrolyzed in boiling
zation at the Western Regional Re- hydrochloric acid solution to yield an
search Laboratory. insoluble crystalline product, tomati-
We have investigated the antibiotic dine hydrochloride, and a clear super-
principles in the tomato plant. The natant solution rich in reducing sugars.
total substance (or substances) in Tomatine melts at 263° to 267° C.
crude tomato plant extracts that ex- with decomposition.
hibited antibiotic activity was named Tomatidine hydrochloride is easily
tomatin. Tomatin inhibits both gram- converted to tomatidine, which is now^
positive and gram-negative bacteria, considered to be a steroid secondary
but most noteworthy is its ability to amine. In cooperation with the Na-
inhibit certain fungi that arc patho- tional Institutes of ilealth, tomatidine
genic to plants or animals. Included has been chemically degraded to a sterol,
 1950-1951 YEARBOOK OF AGRICULTURE

A16 - û//o -prcgncn-3 ( ß ) -ol- 20-one. not appear until the banana is well-
This compound may prove to be valu- ripened. The antifungal substance,
able starting material in the synthesis which inhibits the growth of disease-
of biologically active sterols. causing fungi, has been separated from
Crystalline tomatine and its agly- the antibacterial fractions.
cone, tomatidine, are more effective
against the pathogenic fungi associated AN ANTIBIOTIC must undergo ex-
with human diseases than against the tensive laboratory and clinical evalua-
fungus, Fusarium oxysporum f. lyco- tion before it can be put into general
persici, that causes the tomato wilt. use as a therapeutic agent. If tests on
They are almost completely without experimental animals show that the
eíTect on the gram-negative bacterium material is either nontoxic or of very
Escherichia coli and are very slightly low toxicity and that it is capable of
effective against the gram-positive protecting or curing infected animals,
bacterium Staphylococcus aureus. The the compound may be tried on humans
oral toxicity of these compounds has with their knowledge and consent.
been determined. Essentially no toxic Care must be taken to prove that the
effects have been observed when albino antibiotic will kill or suppress the
rats were fed diets containing up to growth of a variety of pathogenic mi-
0.04 percent of these two compounds cro-organisms, that it will be readily
for 200 days; likewise, subacute and absorbed into the body, that it has no
acute oral toxicity results were favor- damaging action on body cells, and that
able. However, they are very toxic if it is stable and effective in the presence
administered by intravenous injection. of body fluids, cells, and tissue en-
We have found that sweetpotato zymes. If a compound passes those
vines, which are sometimes used as tests, it may be classified as a satis-
silage, contain highly active antifungal factory chemotherapeutic agent. None
and antibacterial substances. It is per- of the few antibiotics isolated thus far
haps significant that the edible tuber from plants have passed all the tests,
also contains these substances. From although some show promise.
an active water-soluble resinous frac- The search for antibiotics in higher
tion, a buff-colored crystalline-appear- plants is relatively new. Thousands of
ing solid and a clear red-brown liquid, plants have been tested for antibiotic
with a distinctively characteristic odor, activity by numerous investigators, but
have been obtained. The solid rnaterial the search, of necessity, has been nar-
exhibits selective activity toward the row. An investigator docs not have the
gram-negative bacterium {E. coli), and time to consider each plant individ-
the liquid toward gram-positive bac- ually when he is trying to survey a
teria (especially Mycohacteria) and larger part of the plant kingdom. He
toward fungi. first makes a preliminary survey of a
The banana skin has been referred number of plants as efficiently as he
to as "nature's bacteria-proof wrap- can by assaying the extracts against a
per." In our investigations we found sufficient number of representative
that the green banana skin and pulp micro-organisms. From such a survey
contain antifungal substances, but he can select the species and varieties
ripe banana skin and pulp (naturally that warrant his further attention. The
and ethylene-ripcned) contain both problem then is to isolate pure sub-
antifungal and antibacterial sub- stances from the varieties of plants that
stances. Our results indicate that anti- showed promise in the preliminary sur-
biotics in the banana skin and pulp vey and to prove the therapeutic pos-
appear during the ripening process. Of sibilities of these pure substances.
particular significance is the fact that
an antibacterial factor active toward P. S. ScHAFFER^ as biochcmist in
acid-fast bacteria {Mycohacteria) does charge of the antibiotic section^
ANTIBIOTICS THAT GOME FROM PLANTS
733
biologically active compounds division, keting Administration, concentrating
Bureau of Agricultural and Industrial mainly on vitamins, proteins, and
Chemistry, is engaged in research on amino acids. He has also conducted
the detection, isolation, and charac- investigations on equipment arid ap-
terization of antibiotics from agricul- paratus designed by industricd con-
tural sources. He entered the Depart- cerns for research on cereal products.
ment of Agriculture in 1930 and before THOMAS D. FONTAINE^ head of the
his present assignment worked on in- biological active compounds division,
secticides with the Bureau of Entomol- joined the staff of the Bureau of Agri-
ogy and Plant Quarantine and on milk cultural and Industrial Chemistry as a
proteins, fats, and other milk products chemist in the oil, fat, and protein di-
with the Bureau of Dairy Industry. vision of the Southern Regional Re-
WILLIAM E. SCOTT^ an associate search Laboratory in 1941. Before en-
chemist in the biologically active tering the Department of Agriculture,
compounds division. Bureau of Agri- he worked at Mellon histitute on the
cultural and Industrial Chemistry, is proteins, amino acids, and enzymes of
engaged in research on aiitibiotics cottonseed. His fields of research in-
from plants. Before joining the Bureau clude plant disease, antibiotics from
staff he conducted research on food plants, and natural and synthetic plant
products in the Production and Mar- growth regulators.

INTEREST IN the use of drugs to treat rheumatoid arthritis and related diseases
is keen. The most pressing problem in 1950 was to produce cortisone, prcgneno-
lone, artisonc, and desoxycorticosterone in quantities large enough for the needs
of arthritics. The drugs were first prepared from animal sources, but those
sources, it quickly became apparent, yielded too little to meet the demand.
Turning to the plant kingdom, scientists began an intensive search for plants
that contain suitable antiarthritic precursors. They have found that the
sapogenins. a little-known group of compounds, are excellent precursors for
antiarthritic drugs. In plants, the sapogenins are combined with sugars. The com-
bination, known as saponins, is highly poisonous. An acid treatment removes the
sugars from the saponins, after which the nonpoisonous sapogenins can be recov-
ered. Some excellent sources of sapogenins have been found in the yucca, agave,
and yam, which are native to Mexico and our Southwestern States.
Department of Agriculture botanists are collecting and identifying plants
belonging to species that contain sapogenin. Department chemists are investi-
gating the plants to determine whether they contain suitable antiarthritic pre-
cursors. When the best plant sources are known, agronomists will determine the
best growing conditions, geneticists wdll find strains with higher sapogenin con-
tent, and chemists wall develop methods for large-scale isolation of the sapo-
genins and their conversion to antiarthritics.—Monroe E. Wall, Eastern Regioncd
Research Laboratory.