Method for evaluating broth culture media: application to Haemophilus

(1)

0095-1137/78/0008-0520$02.00/0

Method

for

Evaluating

Broth Culture Media:

Application

to

Haemophilus

A. W. BRINKLEY'* ANDT. W. HUBER2

Department of Pathology,TheUniversity ofTexasHealth ScienceCenter,SanAntonio,Texas78284,' and

Houston

Health

Department Laboratory,

Houston,Texas

770302

Received forpublication1August1978

Amethodwasdevisedto testthegrowth-promoting ability ofabrothmedium.

The "dilutetoextinction" method determinesthe inoculumrequiredtodevelop

heavy turbidityinabrothwithovernightincubation. A statistical method

using

Poissondistributionwasusedtoshow thatasingle Haemophiluscellcandevelop

heavy turbidityinanoptimalbroth. The dilutetoextinctionmethodwasusedto

evaluate the shelf life of storedmedia, totitratethe growthfactorrequirements

ofHaemophilus, and to evaluate the use of purified hemin and nicotinamide

adenine dinucleotide inabroth medium for thegrowth ofHaemophilus. Ofthe

media tested, themostsuitable formulation was Mueller-Hinton broth

supple-mented with10,ugof hemin and 10,ugof nicotinamide adenine dinucleotideper

ml. The dilute to extinction method appears to be especially useful in the

developmentof broth mediafor fastidiousorganisms. The method could alsobe

usedtoassurethequalityof other broth mediawhicharerequiredtosupport the

growth of small inocula in the clinicalorresearchlaboratory.

The emergenceofantibiotic-resistant strains

necessitatesantimicrobial susceptibility testing ofHaemophilusinfluenza (1, 14). Several

sus-ceptibility testing procedures have been pro-posed,and avarietyof media have beenused(2, 4, 12, 13, 15-18, 21,23-25).Evaluation ofmedia todetermine the size ofinoculum required for growth hasnotbeenreported. Inoculum size is recognizedas asignificantfactorinsusceptibility testing (3,4, 8, 18, 19, 21,23).Amedium which willnotsupportthegrowthof asmallinoculum

is of questionable value. Large-inoculum

re-quirements as well as failure ofsome Haemo-philus isolates to growin supplemented broth

mediapromptedus todevise a method toeasily quantitate theefficacyof abroth medium.

MATERIALS AND METHODS

Organisms.Atotal of111clinical isolates of

Hae-mophilus, consistingof93H. influenzae, 6 H. para-influenzae,5H.haemolyticus, and7H. parahaemo-lyticus,wereused in theevaluation ofthe broth

me-dium. Identification tospecies wasbasedon growth factorrequirementsasdetermined by X, V, andXV

disks (Difco) onMueller-Hinton (MH) agar and by hemolysis on rabbit blood agar.

Media. MH-hemin-nicotinamideadenine

dinucle-otide (NAD)wascomposed of MH broth+10 ug of hemin per ml+10,ugof NAD per ml. Chocolate agar consisting ofhemoglobin (5%,Difco) andsupplement

B (1% Difco) added totryptic soy agar (Difco) was

used to maintain stock cultures and for plate count

determinations.

Hemin stock solution. A 50-mgamountof hemin

(SigmaChemicalCo.)wasmixed with 12.6 mlof0.5

MNa2HPO4,heatedtodissolve,and addedto86mlof

distilled waterand 1.6ml of1 MKH2PO4 (5). This solutionwasautoclavedat121°Cfor15min. Thefinal concentrationwas 500,ug/mi.Thesterile solutionwas

storedat4°C.

NAD stock solution. ,B-NAD (Sigma Chemical

Co.)wasdissolved in distilledwater at aconcentration of 5 mg/ml and sterilized by filtration. The sterile solutionwasstoredat4°C.

Evaluation of basicgrowth media. Four base media, brain heart infusion (Difco), Trypticase soy broth (BBL), 1% tryptose (Difco), and MH broth (Difco), were evaluated for theirabiity to support growth of Haemophilus species. Each medium was supplemented with 1%supplementB(Difco)tosupply X and V factors and tested with and without 0.1% agar. Nine strains ofHaemophilus were cultured on chocolate agar and then suspended in saline to give a turbidsolution.Adrop of each suspension was inocu-lated into5ml of each mediumintriplicate. Growth

was determined by visual turbidity after overnight incubationat35°C.

Evaluation of media by "dilute to extinction"

technique. Tenfold serial dilutions through 10-' of

anovernight broth culture of Haemophilus were made

in eachmedium. The totalviable count of the inocu-lumculture was determined by spread plates, using

chocolateagar. Thedilutiontubes and spread plates

wereincubated at35°Cfor24 h. Media interpreted as yielding heavy growth showedeasilydiscernible tur-bidity corresponding to approximately 108organisms

per ml.The minimum number of cells per milliliter required to yield heavy growth was calculated by

520

on February 7, 2020 by guest

http://jcm.asm.org/

(2)

multiplying the total viable count by the highest di-lution.

Titration of hemin requirement. The effect of hemin concentration on the growth ofsmall inocula was evaluated by the dilute to extinction technique. Concentrations of hemin from 0.001to10,ug/mlwere tested in MH + 20

gg

of NAD perml. Four

Haemo-philus strains were tested.

Titration of NAD requirement. Concentrations of NAD from 0.0001 to10,ug/mI were addedtoMH + 10 ,ug of hemin perml. The NAD requirement for four strains ofHaemophilus was determined by the dilute to extinction method.

Poissondistribution. An overnight culture of H. influenzae H125repeatedly yielded 1.2 x 108 orga-nisms per ml. Based on this observation, dilutions were madeto yield a final dilution containing approx-imately onecell permilliliter. (The starting population was concurrently enumerated by the conventional spread plate technique). Fifty 1-mlsamplesfrom the dilution estimated to contain one cell per milliliter were transferred to tubes containing 4 ml of sterile MH-hemin-NAD. The tubes were shaken, incubated overnight at35°C, and observed for growth. The ob-servednumber of "no growth" tubes was comparedto the expected number of no growth tubes estimated by the Poissondistribution (22):

P(x)=

e-where x is the number of cells in a 1-ml sample, e is the base of thenaturallogarithm, tis the true number

ofcells permilliliterin the test solution (calculated from theplate count of the starting population), and P is theprobabilitythat a1-ml samplewillcontain x cells.

RESULTS

Selection of basic growth medium. The

results shown in Table 1 indicate that brain

heart infusion and MH were superior to the

other media. However, without the beneficial

effect of agar, MH wassuperior tobrain heart

infusion andwaschosenasthe base for further evaluation.

Evaluation of mediabythe dilute to

ex-tinction

method. Theshelf life ofMH+

sup-plement B withand without 0.1% agarwas

eval-uatedbythedilutetoextinctiontechnique.

Fig-ure1

depicts

the beneficialeffect ofaddingagar to

supplement

B-enriched MH. The medium deteriorateddespite the addition of agar. Inan

attempttofindthe causeofdeteriorationupon

storage, additional supplementB(1%),20,ugof

heminperml, and 20,ug of NAD per mlwere added toseparateportionsof7-day-old MH +

supplement

B.

Only

the additionofhemin

re-duced the

required

inoculumtos10

organisms

per ml in the

aged

MH +

supplement

B.

Note,

however,that MH

plus

purified

hemin and NAD

(MH-hemin-NAD) requires

S10

organisms

per

TABLE 1. Comparison of basic media for the growthofHaemophilus spp.a

No.

Medium positive'/no.

tested

BHI+supplement B.. 4/9

BHI +supplement B + 0.1% agar. 9/9

TSB +supplement B ... .. 4/9

TSB+supplement B + 0.1% agar ... 7/9 Tryptose +

supplement

B.. 3/9

Tryptose+supplementB + 0.1% agar.. 7/9

MH+supplementB ... 8/9

MH+supplement B + 0.1% agar 9/9

a H.

aegyptius,

2; H.haemolyticus,1; H. influenza, 2;H. parahaemolyticus, 2; H. parainfluenzae, 2. BHI, Brainheartinfusion; TSB, Trypticase soy broth.

b Visual turbidity resulting from growth of an inoc-ulum of approximately 5x 104organismsperml.

DAYS

FIG. 1. Effectofstoring media (4°C, O to 7days)

on the size of inoculum ofH. influenza H125

re-quired to develop heavy turbidityovernight. Media tested: MH-supplement B (O), MH-supplement B with 0.1% agar(+), and MH-hemin-NAD(-).

ml to yield heavy growth even after 7 days of

storage.

Theevaluationof severalcommon sourcesof

Haemophilus

growth

factors is shown in Table

2. The

ability

ofmediawith Fildes enrichment or

purified

hemin and NAD to support the growth ofasmall inoculum

proved

either

enrich-ment efficacious. The addition of the

purified

factors did not significantly alter the color or

clarity of the medium. The ease of

detecting

turbidityin theclearermediumwasthebasis for choosingpurifiedfactorsoverFildes enrichment.

Table3 shows the results of thetitration for hemin

requirements

of four isolates of

Hae-mophilus,

usingthedilutetoextinction method.

Wedefinedthe end

point

asthesmallestamount

on February 7, 2020 by guest

http://jcm.asm.org/

(3)

522

of heminwhich would support the growth of an

inoculum of l10 colony-forming units/mi. H.

influenza H10 exhibited thelargesthemin

re-quirement, 10ltg/ml. Similarly, the

concentra-tionofNADrequiredby each strain was

deter-mined (Table4). These titrations indicate that

10

gg

of hemin and 10 ,ug of NAD perml should

be adequate for the growth of Haemophilus provided the hemin does not deteriorate upon

storage. Thelowestconcentration ofhemin (0.1

,ug/ml) required by H. influenza H125 was

added to MH + 10 ,g of NAD per ml and stored

for 7 days. This medium still supported -10

colony-formingunits/mIwhentestedbythe

di-TABLE 2. Evaluationofsomegrowth factorsources

bythe dilutetoextinction method

Growth ofH.influenzaeH125 MH brothplus: from aninoculum

(CFU/ml)'

of: 100 10' 102 103 104

Supplement B (1%, - - - - +

Difco)

Supplement C (1%, - - - + +

Difco)

Fildesenrichment(10%, + + + + +

Difco)

Hemin (10 yg/ml) + + + + + +

NAD(10,ug/ml)

aCFU, Colony-forming

units.

TABLE 3. Titrationofheminrequirementin MH+

NAD(10

pg/ml)

by the dilutetoextinction method

Growth ofs10-CFU/mlofa: Hemin(,ug/ml) H.influenza H.

parain-fluenzae

H10 H13 H125 H81

0.001 - - _ +

0.01 - - - +

0.1 - + + +

1.0 - + + +

10.0 + + + +

aCFU,Colony-forming

units.

TABLE 4. TitrationofNADrequirementinMH+ hemin(10pg/ml)bythe diluteto extinctionmethod

Growth ofs10-CFU/mlofa:

NAD(,ug/ml) H.influenza H. parain-fluenzae

H10 H13 H125 H81

0.0001 - - -

-0.001 - - - _

0.01 + + +

-0.1 + + + _

1.0 + + + +

10 + + + +

aCFU,

Colony-forming units.

lute to extinction method with strain H125.

MH-hemin-NAD broth stored up to 5 weeks

alsosupportedgrowthof c10cells per ml. These results demonstrate the stability of the medium

afterprolongedstorage.

Growth of Haemophilus in

MH-he-min-NAD. One hundred eleven

Haemophilus

isolatesweresubcultured from chocolateagarto MH-hemin-NADand observed forgrowthafter

overnightincubation.Growthwasobservedwith

allorganisms except onestrain of H. haemoly-ticus. Most cultures developed uniform

turbid-ity, although a few organisms grew in clumps whichweredifficulttodisperse.

MH-hemin-NAD was evaluated with 13 H.

influenza and 1H.parainfluenzae cultures by

thedilutetoextinctiontechnique.Ail 14cultures yieldedheavygrowthin the dilution tube

con-taining

s10

cells per ml. The highest dilution yielding growthshowed heavy growth. The

di-lute to extinction technique should result in

some tubesreceiving very fewcells, suggesting

that asingle cell could yieldheavy growth with

24h ofincubationin MH-hemin-NAD.

Poissondistribution.ThePoissonequation,

which predicts the distribution of particles in

dilutesolutions,wasusedto testthe

hypothesis

thata single cell would yield heavy growth in

MH-hemin-NAD.

The totalviablecountoftheoriginal

popula-tion was determined by the spreadplate

tech-nique to be 7.8 x

107

organisms per ml. The

dilution blanktested

(10-80792)

contained,onthe

average, 0.65organismper ml.Solving the

Pois-son equation for x =0, P(0) = e-, where ,u =

0.65, wefindthattheprobability ofhavingzero

cells in a sample is 0.52; i.e., 52% of the 50

samples inoculatedshould have received nocells

and result in no growth. The distribution

ob-served resulted in 54% nogrowth. Solving for x

= 1, P(1) = 0.34;

i.e.,

34% of the 50 samples

should have received one cell. If two or more

cellsarerequired for growth,then the

probabil-ity ofnogrowth wouldbeP(0)+ P(1), or 86% of

the 50 samples. Due to the close agreement

between observed and calculated results, it

ap-pears thatonly one organism is required to yield

heavy growth inMH-hemin-NAD broth.

DISCUSSION

MH medium is unique among the fourmedia

tested in that it contains no peptone. Certain

peptones inhibit H. influenza, and reducing

agents such asdithionate and sodium oleate may

neutralize

potential

inhibitors (9, 11). Dubos

iso-lated a fractionfrompeptone that was

bacteri-ostaticforpneumococci only when oxidized and

concluded that the redox

potential

of the

on February 7, 2020 by guest

http://jcm.asm.org/

(4)

dium was critical for the initiation of growth

from small inocula (7). Addition of 0.1% agar

had a marked effect on the growth of

Hae-mophilus inpeptone-containingmedia butlittle

effectwithMH.Starch, present in MH broth as

a"protective colloid" (20), has an effect similar

tothat of agar on the growth of other pathogens

(6). Agar or starch may delay oxidation,

neu-tralize inhibitors, and/or localize extracellular

enzymes necessary forgrowth.

The dilute to extinction technique allows a

quantitative evaluation of broth media and is

based on the premise that an ideal medium

should allow the rapid growth of small

inocula-evenone cell. Amedium is evaluated

by determining the minimum number of

orga-nisms per milliliter required to produce heavy growth overnight; the smaller the inoculum

re-quired, the better the broth.Thus, this dilute to

extinction method notonlycan be used to

com-pare differentmedia as demonstrated here, but

alsocouldbeusedinthe quality controlof broth

media in the clinical or research laboratory

wherethe growth of small inocula is critical.

The Poissonequationwhichpredicts the

dis-tribution ofparticlesindilute solutions was used

to test the hypothesis that a single cell would

yield heavy growth inMH-hemin-NAD in 24 h.

Applicationofthe Poisson equation is an

excel-lentway toevaluatetheabilityof amedium to

support growth. In our earlyattempts to

dem-onstrate aPoissondistribution, 1-ml

samples

of

the dilutedculture wereadded tosterile tubes.

Undertheseconditions inconclusive resultswere

obtained. Parallel Poisson distribution tests were performed with 1-ml samples added to

sterile tubes and 1-ml samples addedto 4ml of

sterile medium. Growth in the 5-ml cultures

demonstrated Poissondistribution,whereasthe

1-ml cultures yielded considerably more no

growth tubes than expected. Perhaps increased oxidation inhibited growth in the smaller

vol-ume.The useof the

larger

volume is

suggested

when maximumrecoveryis

critical,

suchasthe

culture of

cerebrospinal

fluid.

Other

investigators

have

reported

the

growth

and/or antimicrobial

susceptibility testing

of

Haemophilus in brain heart infusion

supple-mentedwithheminand NAD (10, 13,24). The minimuminoculum

required

to

produce

growth

was not reported, and relatively large inocula

were used. The "inoculum effect" is a major

problemin susceptibility

testing

of

Haemophi-lus (4, 8, 18, 19, 21,

23).

In

general,

as the

inoculum increases, the minimum

inhibitory

concentration increases.

Thornsberry

and

Kirven (23) noted a 256-fold increase in the

ampicillinminimuminhibitory concentrationof

a susceptible strain ofH. influenzawhen the inoculum was increased from 105 to 106

orga-nismsperml.Emersonetal. (8) reported a

512-fold increaseinthe ampicillin minimum

inhibi-toryconcentration of resistant strains of H.

in-fluenzaeasthe inoculum increasedfrom 103 to

106 organisms per ml. Roberts et al. (21) and Bottone et al. (4) attribute these inoculum

ef-fectsto the inductionand growth of cell wall-defectiveHaemophilus at the higher

concentra-tions ofbacteria. Roberts et al. (21) have shown

thatcell wall-defectiveHaemophilus possess

un-usual osmoticstability. Asmall inoculum is

re-quired for the susceptibility testing of

Hae-mophilus in order to eliminate the inoculum

effect. Theability of MH-hemin-NAD to

sup-port asmall inoculumseems to recommend its

usein antimicrobialsusceptibilitytesting. The possibility of using MH-hemin-NAD

broth for susceptibility testing was evaluated

further. Tube dilution susceptibility tests for

Staphylococcus aureus ATCC 25923 in

MH-hemin-NAD were compared with the results

obtainedinTrypticasesoy broth. The

antibiot-ics testedwereampicillin, chloramphenicol,

clin-damycin, erythromycin,gentamicin, kanamycin,

cephalothin, andtetracycline. Minimum inhibi-toryand bactericidal concentration resultswere identical in most cases, varied by one dilution in afewinstances, and

only

onceshowedafourfold discrepancy. Moreover, resultswere reproduci-ble when five strains of H. influenza were

tested repeatedly. MH-hemin-NAD appeared

to be an excellent medium for the growth and

susceptibility testing of Haemophilus. Other media are probably adequate, but their ability

to support small inocula should be tested and

considered animportantparameter.

L1TERATURE CITED

1. AmericanAcademyofPediatrics Committeeon In-fectious Diseases.1975.Ampicillin-resistantstrainsof Hemophilus influenzatype B.Pediatrics55:145-146. 2. Barker, J.1974.The in vitrosensitivity ofHaemophilus influenza to clindamycin. Med. Lab. Technol. 31:167-174.

3. Bauer,A.W.,W.M. M.Kirby,J.C.Sherris,andM. Turck. 1966. Antibiotic susceptibility testing by a standardized single disk method. Am. J. Clin. Pathol. 45:493-496.

4. Bottone, E.G.,Z.Brandman,and S. S.Schneierson. 1976.SpheroplastsofHaemophilusinfluenzainduced bycellwall-activeantibioticsand theireffectupon the interpretation of susceptibility tests. Antimicrob. Agents.Chemother.9:327-333.

5. Butler, L.O. 1962. Adefined mediumforHaemophilus

influenzaandHaemophilusparainfluenzae.J.Gen. Microbiol.27:51-60.

6. Casman,E. P. 1947. Anoninfusion bloodagar base for neisseriae,pneumococciandstreptococci.Am.J.Clin. Pathol.17:281.

7. Dubos,T. 1930. Thebacteriostaticaction of certain

on February 7, 2020 by guest

http://jcm.asm.org/

(5)

ponents ofcommercial peptones asaffected by condi-tions of oxidation and reduction. J. Exp. Med. 52:331-345.

8. Emerson, B. B., A.L. Smith, A. L.Harding, and D. H. Smith. 1975. Hemophilusinfluenza type B suscepti-bilityto17antibiotics. J. Pediatr. 86:617-620. 9. Evans, N.M., and D. D. Smith. 1974. The inhibition of

Haemophilusinfluenzae by certain agar and peptone preparations.J.Med.Microbiol.7:305-310.

10.Farrar, W. E., Jr.,and N. M. O'Dell.1974. Beta-lacta-mase activity in ampicillresistant Haemophilus in-fluenzae. Antimicrob. Agents Chemother. 6:625-629. 11.Gilder, H., and S. Granick. 1947. Studies on the He-mophilus group of organisms. J. Gen. Physiol. 31:103-117.

12.Gordon, R. C., T. R. Thompson, L. I. Stevens, and W. H.Carlson. 1974. In vitrosusceptibility of Haemophi-lusinfluenzaetoeight antibiotics. Antimicrob. Agents Chemother. 6:114-115.

13.Gray, B. M., C. A. Hubbell, and H. C.Dillon, Jr. 1977. Manner and meaning ofsusceptibilitytestingof ampi-cillin-resistant Haemophilus influenza. Antimicrob. AgentsChemother. 11:1021-1026.

14.InfectiousDiseaseand Immunization Committee of theCanadianPaediatricSociety. 1975. Ampicillin-resistant strainsof Hemophilusinfluenzae. Can. Med. Assoc.J. 113:222.

15. Jorgensen,J. H., and P. M. Jones. 1975. SimplifSed medium for ampicillin susceptibility testing of Haemophilus influenza. Antimicrob. Agents Chemo-ther. 7:186-190.

16.Jorgensen, J.H., and J. C. Lee. 1977. Standardized disk-diffusionsusceptibilitytest forHaemophilus influ-enzae.Am.J. Cin. Pathol. 67:264-268.

17. Kammer, R. B., D. A. Preston, J. R. Turner, and L. C.Hawley. 1975.Rapiddetection of ampicillin-resist-antHaemophilusinfluenza and theirsusceptibility to sixteen antibiotics. Antimicrob. Agents Chemother. 8:91-94.

18. McLinn,S. E., J. D.Nelson,and K.C. Haltalin.1970. Antimicrobialsusceptibility of Hemophilusinfluenza. Pediatrics 45:827-838.

19. Marks,M.I., and G. Weinmaster.1975.Influencesof media and inocula on the in vitro susceptibility of Haemophilusinfluenzatoco-trimoxazole, ampicillin, penicillin, and chloramphenicol. Antimicrob. Agents Chemother. 8:657-663.

20. Mueller, J. H., and J. Hinton. 1941. A protein-free medium for primary isolation of the gonococcus and meningococcus. Proc.Soc. Exp. Biol. Med.48:330-333. 21. Roberts, D. E., A.Ingold,S.V.Want,and J. R. May. 1974.Osmotically stable L forms of Haemophilus

influ-enzae and their significance in testing sensitivity to penicillins.J.Clin. Pathol.27:560-564.

22. Snedecor, G.W., and W. G. Cochran.1967.Statistical methods, 6th ed., p.223.IowaState University Press, Ames.

23. Thornsberry, C., and L. A. Kirven.1974.Antimicrobial susceptibility ofHaemophilus influenza.Antimicrob. AgentsChemother.6:620-624.

24. Vega, R., H. L. Sadoff, and M. J. Patterson. 1976. Mechanisms of ampicillin resistanceinHaemophilus influenza type B. Antimicrob. Agents Chemother. 9:164-168.

25. Washington, J.A., II, R. J. Snyder, and P. C.Kohner. 1976. Spurious ampicillin resistance by testing Haemophilusinfluenzawithagar containing supple-ment C. Antimicrob.Agents Chemother. 9:199-200.