THE EFFICIENCY OF CERTAIN GERMICIDES IN THEPRESERVATION OF BIOLOGIC PRODUCTS,

W. G. MALCOLMFrom the Department of Bacteriology and Pathology, University of Maryland

Received for publication, June 6, 1931

The matter of preservation is of great importance in the manu-facture of biologic products that are to be employed in the preven-tion and treatment of disease. Phenol and Trikresol are generallyused for this purpose, but they are not entirely satisfactory.With the comparatively recent development of a number of newgermicidal substances, it seemed appropriate to make a study ofsome of them to determine whether or not they might be employedto advantage.There are five features of major importance which must be

known in order to judge the potentialities of a germicide as abiologic preservative. These are:

1. The germicidal action upon spores and vegetative cells of bacteriain different media.

2. The effect upon the antigenic or immunizing value of the variousproducts.

3. Effect upon the appearance of the product and the precipitationof protein or other substances.

4. Toxicity of the germicide, with special reference to the margin ofsafety as measured by intravenous, intramuscular, and intra-spinal toxicity tests.

5. The histologic changes which the germicide produces in animaltissues.

The problem was divided into two parts with different methodsof approach. The first part was a study of the germicidal power

1 A part of a thesis submitted to the Faculty of the Graduate School of theUniversity of Maryland in partial fulfilment of the requirements for the degree ofDoctor of Philosophy.

403

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

of the various chemical substances, together with the physicaleffect upon the basic media employed. Only substances foundsatisfactory in the first phase of the problem were given furtherstudy. The second part of the work was a study of the toxicityof the preservative, its effect upon the appearance and potencyof the biologic products, and the histologic changes it producesin the tissues of experimental animals.The following germicides were chosen:

1. Phenol (crystals) U. S. P.2. Hexylresorcinol (S.T. 37) (1: 1,000)3. Colloidal chloro-cresol (25 per cent)4. Colloidal chloro-thymol (25 per cent)5. Colloidal chloro-cresol (50 per cent)6. Colloidal chloro-thymol (50 per cent)7. Trikresol8. Yatren (powder)9. Mercurophen (2 per cent aqueous solution)

10. Metaphen (2 per cent in 2 per cent NaOH solution)

I. GERMICIDAL TESTS. DETERMINATION OF GERMICIDAL POWER IN

SALT SOLUTION, BOUILLON, AND SERUM, TOGETHER WITH COL-ORATION AND PRECIPITATION OF PROTEIN SUBSTANCES

As menstruums, sterile isotonic salt solution and plain bouillonmade from beef infusion to which 1 per cent peptone and 0.5 percent sodium chloride had been added were used. The reactionwas adjusted to a pH value of 6.8. The medium was checked forsterility after having been autoclaved for one-half hour at 15pounds pressure. Sterile horse serum was also employed.Method of making dilutions. The dilutions were made so that

the total volume of chemical and medium was 5 cc. The germi-cide was added directly to the medium without previous dilutionwith distilled water, because it was found in earlier experimentsthat otherwise the medium was diluted to such an extent that itno longer supported growth.

Organisms employed. Four bacterial species were employed ineach dilution of the germicide; namely, Escherichia coli, Staphylo-coccus albus, Bacillus mycoides, and Bacillus subtilis, because it

AAA

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

has been demonstrated that many germicides are more effectiveagainst some species of bacteria than others. The first twospecies produce vegetative cells only and the latter two are spore-forming bacteria.Inoculations and controls. One-tenth of a cubic centimeter of

each of the cultures was added to 10 cc. plain broth and the wholeincubated for twenty-four hours at 37.5°C. At the expiration ofthis period the cultures were filtered through sterile filter paperto remove the clumps and to produce as uniform a suspension ofthe bacteria as possible. One-tenth of a cubic centimeter of eachof these cultures was added to the specific dilutions of the germi-cide, using one tube for each dilution for each of the four organ-isms. After inoculating the various dilutions, they were allowedto stand at room temperature for twenty-four hours.

Determining the germicidal action of the chemicals. The tubeswere next examined for visible growth and the results were re-corded. Then each dilution was plated separately on plain agar.One cubic centimeter of each of the dilutions was seeded over thesurface of plates that had previously been poured and incubatedfor sterility. The inoculated plates were then placed at 37°C.for twenty-four hours and examined for bacterial growth.

Physical effect of the germicides upon the medium. Observationswere made (a) upon the coloring action of each germicide on thevarious media employed, (b) the appearance of turbidity in thesalt solution, (c) precipitation of protein or other material in thebouillon and serum and (d) miscibility of the germicide with themedium.

Study of germicidal actionPhenol, due to its antiseptic properties, has long been employed

as a preservative for biologic products. Phenol, 1:333 parts inisotonic salt solution, 1: 200 parts in bouillon, and 1: 200 parts inserum killed all four species of bacteria within the twenty-fourhour observation period and caused no discoloration of the saltsolution, bouillon, or serum. There was precipitation in thebouillon and serum at the dilutions in which it was germicidal.This precipitate formed almost immediately and increased with

405

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

agitation. In a comparatively short time it settled to the bottomof the tube leaving the medium practically clear. Phenol wasalso found to be miscible with the media at the dilutions employed.

Trikresol (ortho-, meta- and para-cresol). Trikresol gave aclear solution in aqueous 0.9 per cent sodium chloride. It causedno discoloration of the serum or bouillon, but, in the germicidaldilutions, gave a precipitate in both bouillon and serum. Bac-teria were killed in salt solution containing Trikresol in a dilutionof 1:500, in Trikresol-bouillon 1:333 parts, and in Trikresol-serum 1:250 parts.

CoUoidal chloro-thymol (25 per cent). This is an amber-yellowsolution, water-soluble, and with a weak thymol odor. Colloidalchloro-thymol is a chlorinated methyl-isopropyl-phenol of theformula C6H20H(1)CH3(3)Cl(4)CXH7(6). In this preparation,by a substitution of the hydrogen atom, the halogen chlorine isintroduced into the benzene nucleus. Graphically it is:

CHs

Cl0

CH

CH, CH,

It was expected that the introduction of a halogen into the ben-zene nucleus would decrease its toxicity, and would thereby corre-spondingly increase its germicidal value (Laubenheimer, 1928).

This substance was miscible with the medium, gave the saltsolution a cloudy, bluish-white appearance, and was effective ina dilution of 1:333. In a dilution of 1:200 in bouillon and serum,it killed all four test organisms. In both there was marked tur-bidity with heavy precipitation.

Colloidal chloro-cresol (25 per cent). This cresol derivative,also, is an amber-yellow solution, water-soluble, with a weak odorof cresol. In this preparation, as with colloidal chloro-thymol,the hydrogen atom is substituted by the halogen, chlorine.

Colloidal chloro-cresol (25 per cent) in salt solution was effec-tive in a final dilution of 1: 500, giving a pale pearl-gray color to

406

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

the solution. It was effective in a dilution of 1: 333 in serum and1:250 in bouillon. In the bouillon and serum there was heavyprecipitation, with a rapid settling of the precipitate.

Colloidal chloro-cresol (50 per cent). Colloidal chloro-thymol (50per cent). This strength of these substances was recommendedby the manufacturer after failure of the 25 per cent solution toprove satisfactory in this work, due to the heavy precipitation ofthe protein material. Dilutions were made in salt solution,bouillon and serum from 1 per cent to 0.1 per cent. Althoughthere was an increase in the germicidal action over the 25 per centcolloidal chloro-cresol and chloro-thymol, there was also increasedprecipitation.

Ether as a menstruum of preservatives for biologic products iseffective for phenol and Trikresol in preventing precipitation, andso was added to the above colloids in equal parts in both the 25and the 50 per cent strengths.The results in the 50 per cent concentration were the same as

those in the 25 per cent concentration, except in the case of thechloro-thymol. Here, in a concentration of 0.4 per cent it gave aheavy precipitate, and a slight precipitate in 0.2 per cent in broth.In a strength of 0.1 per cent a precipitate occurred in the serum.The chloro-cresol in 0.3 per cent concentration precipitated

bouillon and serum slightly. The results obtained in the precipi-tation of the protein materials, by colloidal chloro-cresol andcolloidal chloro-thymol in both the 25 and 50 per cent strengthsshowed that ether would not prevent precipitation to an extentthat would make their employment practicable.

Yatren. This is a German preparation, the chemical nature ofwhich was not known by the author. It is a brown powderhaving a maimum solubility in water of 4 per cent. It is notgermicidal in a 1:50 dilution. The media were colored from adeep brown to a light amber depending upon the dilution. In a4 per cent solution there was no precipitation in bouillon or serum.

Hexylresorcinol (S.T. 37) (1:1,000). In a 1:1,000 dilutionthe solution is water clear. This substance has the formulaC6Hs(OH)2(C6H13)2 and was miscible with the media. In a con-

2 New and Non-official Remedies, 1929.

JOURNAL OF BACTERIOLOGY, VOL. XX:I, NO. 6

407

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

centration of 1: 50 it was not germicidal in any of the basic media.It caused no discoloration and no precipitation of the bouillonor serum.Metaphen (2 per cent in a 2 per cent NaOH solution). Meta-

phen, a proprietary preparation, known as 4-nitro-3, 5 bisacetoxy-mercuri-2-cresol (Raiziss and Sevarac, 1923). Its structuralformula is:

OHr/CHs

CH3C,O--Hg0HgOOC CH3NO2

The 2 per cent solution is amber colored and it mixes readily withthe media employed. Because of its high germicidal value, itwas necessary to make an initial dilution of 1: 1,000, using dis-tilled water, before preparing the desired dilutions with the media.Metaphen gave the media a faint amber colored appearance,

but one that was not easily discernible. In a concentration ashigh as 1: 200 there was no precipitation in either the bouillon orserum. It was germicidal in salt solution in a dilution of 1: 10,000,in 1: 8,333 parts in bouillon, and in 1: 5,000 parts in serum.For further study of the germicidal action of this substance, a

number of bottles were prepared, some containing serum, andsome bouillon. Metaphen, 1:8,000 in bouillon and 1: 5,000 inserum was employed. Each was inoculated with about 0.5 gramof dust particles. These cultures showed no growth when indirect sunlight for several weeks, after which time there appearedin the serum a growth of air bacteria. At the end of six weeksthe bouillon showed the presence of molds and various organisms.Metaphen in a 1:8,000 concentration was added to 100 cc. oftetanus toxin and was placed in each of five bottles. Also fivebottles containing a similar amount of diphtheria toxin withMetaphen in a 1:8,000 dilution were inoculated with dust andincubated for two weeks at 37°C. At the end of this time theywere examined and none showed growth. From these, fivesamples were placed in the sunlight and five in the dark. At theend of a month all five samples kept from the sun rays were sterile

408

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

while three of the five samples exposed to the action of sunlightshowed growth. While no definite conclusions are drawn as tothe effect of sunlight in causing a loss of the germicidal propertiesof Metaphen, it seems to have some apparent destructive action.Mercurophen (2 per cent solution). Mercurophen, a propri-

etary preparation, is sodium-parahydroxymercuri-ortho-nitrophenolate. Structurally it is as follows (Raiziss and Sevarac,1923):

0 NaONO2

HgOH

The 2 per cent aqueous solution is odorless with a brick-red colorwhile in higher dilutions it is amber. In preparation for thegermicidal tests, due to its ability to kill bacteria in high dilu-tions, a preliminary dilution of 1:1,000 in distilled water wasmade before completing the final dilutions by the addition of thebasic media.

This substance was miscible with the media, and was germicidalin salt solution in a dilution of 1: 16,500; bouillon, 1: 12,500; andserum, 1:8,500. It imparts an amber hue not unlike that ofnormal serum and bouillon, but slightly more intense. Therewas no precipitation in serum in dilutions as low as 1: 200.A series of preparations were made, as in the case of Metaphen,

under identical conditions, and exposed to sunlight to ascertainif there was loss of gercidal power. Bouillon containingMercurophen in 1:12,000 parts, and serum 1:8,000 parts wereprepared. After four months' exposure at room temperature theserum and bouillon were found to be sterile. After inoculatingfive samples of tetanus toxin, and five samples of diphtheria anti-toxin, each containing 1: 10,000 parts Mercurophen and incubat-ing at 37°C. for three weeks, all were found to be sterile. Thesewere then divided into two lots, one exposed to the air and sun-light, and the other exposed to the air but protected from sun-light. At the end of a month both lots were sterile. It was

409

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

assumed from this experiment that light does not affect thegermicidal power of Mercurophen.

DiscussionThe germicidal power refers to the destruction of all the species

of bacteria employed in the experiments. A large inoculum wasintentionally used to provide for exceptionally heavy bacterialgrowth, one that would far surpass the usual conditions of acci-dental contamination. If the dilutions are not made in such amanner as to contain sufficient medium to promote growth ofthe bacteria, the germicidal effectiveness is given a high andfalse value.The germicidal power of each substance decreased with an

increasing amount of organic material present in the mediumemployed. There was also noted a selective action of the chem-ical substances, the organic mercurials being particularly effectiveagainst Staphylococcus albus. The tests were repeated a numberof times and the germicidal and physical effects were regularlyfound to be as stated.Of the chemicals studied those best suited for biologic preserva-

tives from the standpoint of germicidal efficiency, physical effect,and the appearance of the product were Mercurophen, Metaphen,Trikresol, and phenol, in the order given. The remainder of thepaper is concerned with a further study of these substances.

II. TOXICITY TESTS

This phase of the work was undertaken chiefly to determine thetoxicity of Metaphen and Mercurophen, but included in partthe toxicity of phenol and Trikresol. The toxicity of phenol andTrikresol has been determined by Leake and Corbett (1917) anda number of other workers, with comparable results.Leake in 1917 used mice as experimental animals. Assuming

the minimum lethal dose to be the amount necessary to kill 80per cent of the mice at a given dose, Leake found that phenol hasan M.L.D. of 0.00037 gram per gram of mouse. Trikresol has thesame M.L.D. as found for phenol. The toxicity of Trikresol andphenol was not lessened when diluted with normal horse serum.

410

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

MethodsThe experimental animals employed in this part of the problem

were rabbits and albino rats. The injections were made intrave-nously and intramuscularly.The phenol was weighed and diluted to the required concentra-

tion with double distilled water. Trikresol, Mercurophen andMetaphen in the liquid state were diluted to the desired concen-tration with distilled water. The quantity of each germicide tobe injected into each animal was calculated by multiplying theweight of the rat or rabbit by the dose in grams of the germicideand dividing the total by the grams of germicide per cubic centi-meter of solution. In the rat, intravenous injections were madein the saphenous vein, while intramuscular injections were madein the fleshy portion of the hind leg. With rabbits, intravenousinjections were made in the marginal ear veins. The rats werebetween 100 and 150 grams in weight, the weight being takenbefore a heavy feeding. Rabbits were not less than 1.5 kgm. inweight and were apparently in excellent physical condition. Theamounts injected were measured in 2 cc. Luer syringes, graduatedin 0.1 cc., and also a 1 cc. syringe graduated in 0.01 cc. Thequantity to be injected governed the choice of the syringe tobe used.The maximum tolerated dose (M.T.D.) was employed to demon-

strate the tolerance of the experimental animals for the germicidesstudied. This degree of toxicity was determined rather than theminimum lethal dose as it was thought that it would representmore satisfactorily the margin of safety that would be essentialfor the safe employment of each respective germicide in thecapacity of a biologic preservative. In these experiments theM.T.D. was considered to be the maximum amount of germicidenecessary to kill 60 per cent of the test animals within a fourteen-day period. Animals living fourteen days after the time of injec-tion were discharged as survivors.

Dilutions of germicidesToxicity tests on phenol and Trikresol were carried only far

enough to compare their reactions with those produced by Meta-phen and Mercurophen.

411

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

The M.T.D. for both phenol and Trikresol in rats was 150 mgm.per kilogram intravenously, and 250 mgm. intramuscularly.Further tests for the M.T.D. were discontinued because of theextreme tremors manifested with such low dosages as 30 mgm.intravenously and 70 mgm. intramuscularly. Many times thesesymptoms appeared before the injection was completed.

Toxicity of Metaphen and MercurophenIn determining the M.T.D. of these substances a number of

factors were considered. Of first importance was the strain ofrats used in the tests. Albino rats of the strain usually employedin the standardization of arsphenamine at the National Instituteof Health were used exclusively. Parallel experiments carriedout on rats of another strain showed a lower tolerance for thegermicides than is here reported, the difference occasionally beingas much as 40 per cent on identical dosages. Furthermore,during periods of intense heat, animal tolerance for the mercurypreparations when injected intramuscularly or intravenously wasgreatly lessened. Both substances injected intramuscularly indoses of 12 mgm. per kilogram or more produced some irritation,which was manifested for several minutes. It was stated byKolmer (1926) that solubility in the tissue, percentage of mercury,and rate of dissociation of mercury influence toxicity and physio-logical activity. The times of death in the toxicity experimentson animals poisoned by the mercury compounds fell into thegroups as outlined by McNider (1924), depending upon the reac-tion and mode of administration of the mercurial substance. Inthe first group were animals which died usually within two days.The second group would generally survive this period but diewithin seven days, and a final group survived both periods, butdied within fourteen days-manifesting a delayed poisoning.

Discu,ssionIt is evident that the M.T.D. for the mercury compounds is not

a constant value. As shown by the effect of dilution and thetolerance of individual animals, there is a broad range of suscepti-biliity. In some preliminary experiments an individual animal

412

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

TABLE 1

Toxicity charts of Metaphen and Mercurophen

PREPARATION TEST ANIMAL

Metaphen

Albino rat

I Rabbit

Mercurophen Albino rat

oSo

per cent

1.01.01.01.02.02.02.02.00.50.50.50.10.10.1

2.02.02.02.02.02.02.02.02.02.0

1.01.01.01.02.02.02.02.00.50.50.50.10.10.1

INUMBEROF

ANIMALSON EACHDOSE

5555444444455S

4444444544

55554444555555

DOSAG | IOUTE OFDOAG INJE0CTION

mgm.perkgm.16.620.024.028.014.016.620.024.02.64.06.04.06.08.0

6.08.010.012.014.016.0

3.04.05.06.0

16.620.024.028.014.016.620.024.04.06.08.08.010.012.0

Intramuscular

Intramuscular

Intramuscular

Intramuscular

IntramuscularIntramuscularIntramuscularIntramuscularIntravenousIntravenousIntravenousIntravenousIntravenousIntravenous

IntramuscularIntramuscularIntramuscularIntramuscularIntramuscularIntramuscularIntravenousIntravenousIntravenousIntravenous

IntramuscularIntramuscularIntramuscularIntramuscularIntramuscularIntramuscularIntramuscularIntramuscularIntravenousIntravenousIntravenousIntravenousIntravenousIntravenous

413

PERCENTSUR-VIVAL

100100600

10010075

010050

010060

0

100100756025251008025

0

1008060

01008020

08000

10020

0

M.T.D.

mgm.perkgm.

24.0

20.0

2.6

6.0

12.0

4.0

24.0

16.6

4.0

8.0

I-~~~~~~~~~~~~~~~~~~~~~~

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

TABLE 1-Concluded

4 NUMBERPR0~ PROUTE OF CENT T

PREPARATION TEST ANIMAL 5 AIMALS DOSAGE M.T.D.ON EAHINJECTION SUE-O DOSAE VIVAL

mgm. mgm.per cent per per

kcgm. kgmt.2.0 4 6.0 Intramuscular 1002.0 5 8.0 Intramuscular 802.0 5 10.0 Intramuscular 802.0 4 12.0 Intramuscular 75 12.0

Mercurophen Rabbit 2.0 4 14.0 Intramuscular 50-Concluded 2.0 2 16.0 Intramuscular 0

2.0 4 3.0 Intravenous 1002.0 5 4.0 Intravenous 80 4.02.0 4 5.0 Intravenous 252.0 4 6.0 Intravenous 9

would occasionally tolerate a dose as great as 40 mgm. of Meta-phen or Mercurophen per kilogram. The higher concentrationsmay produce thrombosis and cause sudden death. It will benoted in the intravenous experiments in rats that when 0.5 percent solutions were employed, death occurred generally withintwenty-four hours, giving a dosage toleration of 50 per cent lessthan that of the 0.1 per cent solution. From these data it maybe inferred that dilution influences the susceptibility of the ani-mals. This is particularly true when the injection is intravenous.The M.T.D. of Metaphen and Mercurophen in rabbits was found

to be 4 mgm. per kilogram intravenously and 12 mgm. per kilo-gram intramuscularly. The albino rats showed a tolerance tothe mercury compounds of 0.1 per cent solution intravenously and1.0 per cent solution intramuscularly as follows: Metaphen,intravenously 6 mgm. per kilogram; intramuscularly 24 mgm. perkilogram; Mercurophen, 8 mgm. per kilogram intravenously, and24 mgm. per kilogram intramuscularly. Rats showed 100 percent greater tolerance to Mercurophen intravenously for the 0.1per cent solution than for the 0.5 per cent solution. Rabbits hada lower tolerance both for intravenous and intramuscular injec-tions than that exhibited by rats. Phenol and Trikresol areapproximately 25 times less toxic intravenously and 10 times less

414

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

TABLE 2Toxicity charts of phenol and Trikresol

PREPARATION

Phenol

TEST ANIMAL

Albino rat r

Trikresol Albino rat

per cent

2.02.02.02.02.02.02.02.02.02.02.02.02.02.02.02.0

2.02.02.02.02.02.02.02.02.02.02.02.02.02.02.02.0

NUMBEROF

ANIMALSION EACHDOSE

2222233333333333

22222333333333

33

DOSAGE

mgm.perkgm.

10.020.030.040.070.0100.0120.0130.0150.0170.0200.0220.0240.0250.0260.0270.0

10.020.030.040.070.0100.0120.0130.0150.0170.0200.0220.0240.0250.0260.0270.0

ROUTE OFINJECTION

IntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntramuscularIntramuscularIntramuscularIntramuscularIntramuscularIntramuscular

IntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntravenousIntramuscularIntramuscularIntramuscularIntramuscularIntramuscularIntramuscular

415

PERCENTSUR-VIVAL

10010010010010010010066660

100100100100330

100100100100100100100100660

100100666600

M.T.D.

mgm.perkgm.

150

250

150

250

toxic intramuscularly than Metaphen and Mercurophen. How-ever, the M.T.D. is misleading in the case of phenol and Trikresolbecause of the extreme tremors manifested and the "burning" of

k

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

the tissue of the animals in amounts far below the maximumtolerated dosage.

If the toxicity of Metaphen and Mercurophen is placed upona comparative basis, there is little difference between them.Mercurophen appears to be somewhat less toxic than Metaphenwhen used intravenously in rats but both have the same M.T.D.in rabbits.

III. INTRASPINAL TESTS. TO TEST THE ACTION OF METAPHEN AND

MERCUROPHEN AS COMPARED WITH THE ACTION OF PHENOL

AND TRIKRESOL WHEN INJECTED INTRASPINALLY

In 1913, Kramer (1913) observed a number of cases of suddendeath after the treatment of cerebro-spinal fever with intraspinalinjections of specific antiserum, and advanced the hypothesis thatthe preservative, Trikresol, was the cause of the symptoms. Thatthe toxic action of the serum introduced into the subarachnoidspace of dogs and cats was due to this preservative was laterproven by Hale (1913). In 1914, Voegtlin (1914) confirmed theresults of Kramer and of Hale and found that Trikresol and phenolwere equally toxic when injected in like quantities, and stated,"As a general rule, the effect on blood pressure increases in propor-tion to the rate of injection and the concentration of the phenolpreservative."The fact that phenol and Trikresol possess such toxic properties

when injected intraspinally made it desirable to know the effectsproduced by Metaphen and Mercurophen upon the centralnervous system.

ProcedureThe experiments were carried out on dogs. The anesthetic

employed was ether and the degree of anesthesia was kept uniformby performing tracheotomy and introducing the vapor from abottle directly into the trachea. Blood pressure was taken fromthe carotid artery, and tracings were obtained by using a mercurymanometer. Respiratory rates were recorded graphically byinserting a canula into the pleural cavity just above the diaphragmand recordings made on a kymograph connected with a tambour.

416

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

Methods of injectionTwo methods of injection were employed; namely, by pressure

and by gravity. Several vertebrae in the lumbar region wereexposed and injections made by introducing a needle directly intothe spinal canal. Spinal fluid was permitted to escape. Theneedle was then connected with a Luer type of syringe or gradu-ated burette depending upon the method used.Antimeningococcus serum was preserved with Mercurophen

and Metaphen in 1: 2,500 and with phenol and Trikresol in 1: 200concentrations. These dilutions permitted a margin of morethan double the quantity of Metaphen and Mercurophen neededfor preservation purposes. Sera preserved with each of the abovesubstances were injected alternately, using both the pressure andgravity methods of administration.

ProtocolOnly one protocol of several experiments is given since the

others are similar to this one. The symptoms manifested afterthe administration of phenol intraspinally are the same as thoseproduced by Trikresol, so a protocol showing in detail the effectof the latter substance is not included.

Dog 2, 8.2 kgm. Anesthetic, ether.

10:32 a.m., withdrew 1.5 cc. spinal fluid-clear.10:41 to 10:50 a.m., 10 cc. serum preserved with Metaphen 1:2,500

injected intraspinally by gravity method.10: 59 a.m., pressure released and serum permitted to flow out. Blood

pressure and respiration normal.11:10 to 11:12 a.m., 4.1 cc. serum containing 0.5 per cent phenol

administered by gravity method. A drop in blood pressure followed,with shallow and uneven respiration. Immediately before releasinginjection pressure, coarse tremors were manifested. After releasing theserum from the spinal canal, there was an immediate rise in blood pres-sure to normal and respiration became constant.

11:27 to 11:30 a.m., 4 cc. of one part Mercurophen, in 2,500 partsserum administered by gravity method.

11:30 to 11:33 a.m., 1.5 cc. additional serum as above.11:34 a.m., injection pressure released. Blood pressure normal,

417

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

respiration normal. No change in respiration and blood pressure afterthe release of the preserved serums.

11:45 to 11:52 a.m., 4 cc. serum containing Mercurophen (1:2,500)administered by gravity method. Blood pressure and respirationnormal.

12:01 to 12:06 p.m., 4.2 cc. serum containing phenol (0.5 per cent)administered by gravity method. A slight rise in blood pressure wasfollowed by a gradual fall. When injection pressure was released, bloodpressure became normal.

12:10 to 12:11 p.m., 2.5 cc. serum containing phenol (0.5 per cent);rapid decrease in blood pressure, respiration normal.

12:19 to 12:23 p.m., 9 cc. 1 part Metaphen in 2,500 parts seruminjected under pressure. Blood pressure and respiration normal.

12:27 to 12:29 p.m., 2.9 cc. serum containing phenol (0.5 per cent)injected by gravity. Gradual decrease in blood pressure returning tonormal when injection pressure was released.

12:45 to 12:47 p.m., 7 cc. serum containing phenol (0.5 per cent) in-jected under pressure. Marked decrease in blood pressure, respirationbecame shallow and finally stopped; injection pressure was released afterwhich the respiration and blood pressure became normal.

12:54 to 12:58 p.m., normal salt solution containing phenol (0.5 percent) administered by pressure. Marked decrease in blood pressure,tremors were shown, respiration became shallow and finally ceased.

DiscussionPhenol and Trikresol. The results obtained with a serum con-

taining phenol and Trikresol (1:200 parts), were in generalanalogous to those of Voegtlin (1914). In all cases there was animmediate and marked drop in blood pressure with a slowing andfinal arrest of respiration if the pressure was not released. Seruminjected under pressure gave a more precipitous drop than wheninjected by the gravity method. The amount of phenolizedserum necessary to cause these effects was small and withinseveral minutes' time. Respiratory and blood pressure changesare shown in figures 1 and 2.The severe symptoms manifested in the vasomotor and respira-

tory centers with phenol and Trikresol makes the use of thesesubstances as preservatives in biologic products to be administeredvia the spinal canal a hazardous one.

418

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

Metaphen and Mercurophen. A review of the protocols andthe respiratory and blood curves after the injection of these twodiluted mercurial compounds showed an absence of the severesymptoms manifested in the phenol experiments. Injections of

FIGS. 1 AND 2

large amounts of serum preserved with Metaphen and Mercuro-phen had no noticeable effect upon the blood pressure or respira-tory movements as shown in figures 3 and 4. The pressuremethod of injection produced a slight increase in the blood pres-sure but the symptoms shown would indicate that these sub-

419

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

stances were not dangerous. It may, therefore, be assumed thatneither Metaphen nor Mercurophen may be expected to haveany harmful effects upon the vasomotor or respiratory centersin the amounts present in the usual doses of therapeutic sera.

FIGS. 2 AND 3

IV. POTENCY TESTS

Sera without preservative were collected from various manu-facturing laboratories for this phase of the study. Tetanus toxinand antitoxin, and diphtheria toxin and antitoxin were used todetermine the effect of the germicidal substances on their toxicity

420

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

or potency. To the above toxins and antitoxins were added thequantities of each preservative which had been found to begermicidal. The substances used and quantities added were:

Phenol, 0.5 per cent in toxins and antitoxinsTrikresol, 0.4 per cent in toxins and antitoxinsMetaphen, antitoxins 1:5,000, toxins 1:10,000Mercurophen, antitoxins 1:8,000, toxins 1:12,000

These mixtures were then placed in a refrigerator at 4.5°C.together with the controls of unpreserved serum. A preliminarytest was made on these toxins and antitoxins to determine theirexact toxicity or unitage. They were then placed in an incubatorat 37°C. to hasten any action of the preservatives. At intervals,potency tests were made on preserved and unpreserved sera tonote any decrease in potency.

Potency tests. The methods employed in testing the potencyof the biologic products studied were essentially the same asthose used at the National Institute of Health in testing thepotency of standard commercial diphtheria and tetanus toxinsand antitoxins.3The strength of the toxins was measured in terms of M.L.D.'S

per cubic centimeter, being the minimal amount of toxin neces-sary, when injected subcutaneously in the test animals of a givenweight, to produce acute toxic death in approximately ninety-sixhours. The potency of antitoxin was measured in units, a unitbeing the minimal amount of antitoxin which when added to1 L+ dose of toxin and injected subcutaneously into a test animalof given weight would protect the test animal for approximatelyninety-six hours.

DiscussionAs a result of these experments the following conclusions seem

to be warranted:Diphtheria toxin, when preserved with phenol, Trikresol,

Metaphen or Mercurophen, showed no decrease in strength when3Tests of standard and commercial diphtheria and tetanus antitoxin. Pub-

lished by the United States Hygienic Laboratory (Laboratory memorandum).

421

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

compared to that of unpreserved toxin after sixteen and twenty-eight days' incubation at 37°C.

Diphtheria antitoxin similarly preserved showed no loss ofpotency when compared to the potency of the unpreserved serumwhen incubated at 37°C. for sixteen, twenty-two and thirty-nine days.

Incubation at 37°C. for fourteen days caused a complete lossof toxicity in the control and preserved tetanus toxin while tetanusantitoxin preserved as above retained its potency for the periodsstudied, i.e., seven and fourteen days.An analysis of the potency tests showed similar results with no

definite evidence of loss of toxicity or potency due to the actionof any of the chemical substances used as preservatives. Tetanustoxin, however, is unstable at 37°C. and the unpreserved as wellas the preserved product showed a complete loss of toxicity.

V. HISTOPATHOLOGIC STUDY OF THI EFFECTS OF METAPHEN AND

MERCUROPHEN UPON TISSUES

The ability of certain mercury salts to produce nephritis hasbeen known for years.

In 1860 Pavy induced an experimental nephritis with mercury.He considered the Malpighian bodies to be the seat of injury andalso noted necrosis of the tubular epithelium. Later, Aschoff(1912) reported in detail the renal pathology produced by mer-curic chloride, which consisted in swelling, hyaline vacuolation,and necrosis of the tubular epithelium.The present study was undertaken to learn if the mercury com-

pounds, Metaphen and Mercurophen, in doses representing 25per cent of the M.T.D., produced any tissue changes.

Procedure. For this study albino rats were used as the experi-mental animals. One hundred and fifty per cent and 25 per centof the maximum tolerated doses of Metaphen, Mercurophen, andmercuric chloride were administered intramuscularly. Thesedoses were chosen in order that the lesions produced by a toxicdose and sub-tolerated dose might be learned. Mercuric chloridewas also employed so that a comparative study of the lesionsproduced by Metaphen and Mercurophen might be made.

Al0f

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

The animals were killed at varying lengths of time in an attemptto observe the various stages of the lesions produced. Six ratswere used on each mercurial substance in the 150 per cent dosageand six rats on the 25 per cent maximum tolerated dosages. Withthe latter dosage the animals were killed on the fifth, seventh,twelfth, fifteenth, seventeenth, and twenty-first days. Care wastaken to place the tissues in fixing fluids immediately after thedeath of the animal.

Discus8ionThe a ministration of 150 per cent of the maximum tolerated

doses of Metaphen, Mercurophen, and mercuric chloride producedtissue injuries which were essentially identical in character andin degree. Lesions of the heart muscle, spleen, and liver werenot constant, and when demonstrated at all, were of minor impor-tance. The kidney showed the most characteristic changes;namely, a toxic nephrosis.

Mercuric chloride produced necrosis of the muscular tissue atthe site of injection in two animals. This condition was notfound in any of the rats injected with Metaphen or Mercurophen.

It was not possible to differentiate the lesions produced by thethree mercurials studied. The degree of tissue injury produceddepended apparently on the amount of mercury administeredand the length of survival. One-fourth of the M.T.D. producedno evident histologic changes. The quantity of germicide whichwould be present in the maximum dose of serum to be adminis-tered is materially less than one-fourth of its maximum tolerateddose. Therefore, because of the extremely small amount of thetwo mercury compounds, Metaphen and Mercurophen, neededto preserve biologic products, these substances may be used forthis purpose without fear of causing any symptoms or pathologicchanges in persons receiving the maximum therapeutic doses ofthese serums.

SUMMARY

Yatren and Hexylresorcinol (1:1,000) have insufficient germi-cidal power to permit their use for the preservation of biologicproducts.

423

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

W. G. MALCOLM

Colloidal chloro-cresol and colloidal chloro-thymol, either in25 or 50 per cent strengths, are unsuitable as biologic preserva-tives, because of their destructive action in precipitating protein.

In serum, Mercurophen has an approximate germicidal potency40 times greater, and Metaphen 25 times greater, than phenoland Trikresol. Intramuscular injections in rats of 1 per centsolutions of Metaphen and Mercurophen are approximately 10times more toxic than phenol and Trikresol.The higher germicidal power of these two mercury compounds

causes them to be relatively less toxic than phenol and Trikresolwhen used as biologic preservatives, and permits a greater marginof safety in their use.

Neither Metaphen nor Mercurophen at their germicidalstrength precipitates proteins, while both phenol and Trikesolhave this action.The use of phenol and Trikresol in germicidal concentrations

in the spinal canal is dangerous because of action upon the respir-atory and vasomotor centers. Serums preserved with Metaphenand Mercurophen, on the other hand, when injected into thespinal canal had no effect upon the blood pressure or respiratorymovements.

It was demonstrated that the type of tissue injury produced byMetaphen and Mercurophen is entirely dependent upon thequantity of mercury administered. Quantities representing 25per cent of the maximal tolerated dose produce no histologicchanges.As far as could be determined neither phenol, Trikresol, Meta-

phen, nor Mercurophen decreases the potency of toxins andantitoxins.Of the biologic preservatives studied, the mercury compound,

Mercurophen appears to be the most satisfactory, with Meta-phen as second choice, and both superior to phenol and Trikresol.

The author wishes to make record of the kindness of Dr. GeorgeW. McCoy, Director, and members of the staff for the privilegeof carrying out this study at the National Institute of Healthand for their many helpful suggestions.

424

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

PRESERVATION OF BIOLOGIC PRODUCTS

REFERENCESASCHOFF, L. 1912 Zur Morphologie der Nierensekretion unter physiologischen

und pathologischen Bedingungen. Verhandl. d. Deutschen Patho-logische Gesellschaft, 199.

HALE, WORTH 1913 The cause of death from subdural injections. HygienicLaboratory Bulletin, No. 91, P. 5.

KOLMER, JOHN A. 1926 Principles and practice of chemotherapy with specialreference to the specific and general treatment of syphilis. W. B.Saunders Company, Philadelphia, 739.

KRAMER, S. P. 1913 A possible source of danger in the use of antimeningitisserum. Jour. Amer. Med. Assoc., 60, 1348.

LAUBENHEIMER, KURT 1928 Ueber Chlorothymol als Desinfektionsmittel.Aus dem Staatsinstitut fur experimentelle Therapie in Frankfurt, a.M.

LEAKE, JAMES P., AND CORBETT, HUGH B. 1917 Toxicity of certain preserva-tives used in serums, viruses, and vaccines. Hygienic LaboratoryBulletin, No. 110, p. 35.

MAcNIDER, W. DEB. 1924 A review of acute experimental nephritis. Physiol.Rev., 4, 595.

PAVY, F. W. 1860 Poisoning by white precipitate. Guys Hospital Reports,London, p. 504.

RAIZISS, G. W., AND SEVERAC, M. K. 1923 A new organic mercury compoundwith powerful germicidal properties. Jour. Lab. and Clin. Med., 9, 71.

VOEGTLIN, CARL 1914 The pharmacological action of some serum preservatives.Hygienic Laboratory Bulletin, No 96, p. 87.

JOURNAL OF BACTERIOLOGY, VOL. XXI, NO. 6

425

on May 2, 2020 by guest

http://jb.asm.org/

Dow

nloaded from