j. Soc.Cosmet. Chem.,42, 133-145 (May/June1991)

Inhibitorof odor-producing axillary bacterialexoenzymes A. CHARIG, C. FROEBE, A. SIMONE, and E. EIGEN, Colgate Palmolive Research Center,Piscataway, NJ 08854. Received October1O, 1990.

Synopsis

Zinc from aqueouszinc glycinatebindsto hair and skin at pH 7. The bindingincreases with Zn concentration, becomingappreciable whenZn exceeds about 1 millimole/liter. In the presence of soapor synthetic detergents,Zn binding still occurs,but at a lower level. Zinc boundto hair actsas an effectiveinhibitor of two bacterialexoenzymes, aryl sulfataseand betaglucuronidase, which are implicatedin the productionof steroidalaxillarymalodor(1). Clinical testingof the deodorantactivityof this materialhasshownthat a 4.5 % solutionof Zn-GLY is as effectiveas 5% aluminum chlorhydratein controllingaxillary odor. Although zinc glycinate is slightly bactericidal, enzymeinhibition seemsto be the principalmechanismof deodorancy. The potentialfor a new underarm deodorant is discussed.

INTRODUCTION

Axillary apocrinesecretions, asthey emergeat the skin surface,aresterileand odorless; odorarisesfrom the actionof bacteriauponsweat,and a groupof steroids(androstenes) is foundto contributelargelyto it (2). Eigen(3) provideda mechanism for the generation of axillary steroidmalodor:the steroidsare secretedas water-solublenon-volatile conjugatesthat are then cleavedto the volatilefree steroidby esterases, suchas aryl sulfatase (AS) and beta-glucuronidase (beta-G), which are secretedby axillarybacteria (Figure 1). In a previouspaper,we reportedon a seriesof studiesthat providedevidence to supportthis mechanism(1).

Oneof the itemsof evidence supporting our theorywasthe factthat inhibitorsof the enzymesalsofunctionasinhibitorsof odorproductionin vitro.As part of that study,we testedmanypotentialinhibitorsof the enzymes anddiscovered severalcompounds that inhibit them. Such inhibitors, we felt, shouldfunction in the axilla to prevent the generationof odor, ratherthan to maskit, inhibit perspiration,or disturbthe indigenousmicroflora(2-4). A numberof promisinginhibitorswere zinc salts,of limited solubilityin water. We noticedthat activeZn levelsremainedhigher in someof the enzymeassaysand hypothesized that the Tris buffer might be solubilizingzinc by chelation;thus we testedzinc complexedwith severalligands.Among their varying degreesof inhibition, zinc glycinate(Zn-GLY) wasfoundto be effectiveasan inhibitor of both sulfataseandglucuronidase. It is alsoconvenientto prepareand it is stable.We 133

134

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

--OSO2 •O•

c% O'

0"

5.ct-androst16-en-3,[•-ol

5,ct-androst16-en-3,[•-ol sulfate

glucuronide • / O•c•.O vNH2 Zn

C

o9 •. ø/ 'x J Zinc Glycinate

,•-glucuro

sulfatase

HO

3,cz-androst16-en-3.•-ol 3,a-13ydroxystero•d

clehyclrogenase

5,cz-androst- 16-en-3one Figure 1. Inhibitionof enzymes causing axillaryodor.

ODOR

INHIBITION

135

then turned our attentionto the feasibilityof incorporatingthis materialas an active agentinto cleansingformulations.

For an activeagentto exhibitlong-lastingdeodorancy, it must bind to the skin and/or hair, andthe boundresiduemust remaininhibitoryin the presence of soaps,etc. This paperdescribes our discoverythat zinc glycinateis in fact substantiveto skin and hair evenin the presence of a soapor syntheticdetergent(syndet)solution.Oncebound,and in the presenceof thesesolutions,its anti-enzymeactivity is preserved.Thus Zn-GLY may serveas an effectivedeodorant. METHODS MATERIALS

Enzymes,substrates, and buffercomponents werepurchased from SigmaChemicalCo. All experimentswere carriedout in 0.1 M Tris buffer at pH 7.

Non-isotopic Zn-GLYwasprepared asdescribed by DubskyandRabas(6). [65Zn] Zn-GLYwasprepared asneeded foreach experiment. A convenient amount of65Zn,in solutionof negligiblevolume and containingnegligiblemassof Zn, was addedto a solutionof normalZnCI 2 at a concentration twice what waswantedfor the experiment. To it was addedan equal volume of solutioncontaininga two-fold molar excessof glycinesodiumsalt, so the final Zn concentrationwas as required. The mixture had a pH of 9. The precipitate(ZnO), whichformedwhenglycinewasfirst added,gradually redissolved as the complexformed, and the final solutioncontainedthe requiredconcentrationof Zn in solutionas the diglycinateat a pH of approximately8. INSTRUMENTATION

Radioisotope wasdetectedby countingin a Baird Atomic Model 530 Gamma Spectrometer with multichannel analyzer. Fluorescence intensities were measured with a Perkin-Elmer

Model MPF-3

with an

OsramXBO high-pressure Xenon lamp, usingquartz cuvettes.Absorptionwavelength for 4 methylumbelliferol was334 mm; emissionwavelengthwas444 mm. PRELIMINARY

STUDY:

KINETICS

OF BINDING

OF Zn-GLY

TO HAIR

Beforeundertakingquantitativesubstantivitystudies,it wasnecessary to establishthe

timescale ofbinding ofZntothesubstrate. Solutions of[65Zn] Zn-GLYwereprepared as above,with Zn at 0.01 raM, 1 mM, and 10 raM; hair samples(0.1 g each)were exposedto 2.5 ml of solutionfor variouslengthsof time. The solutionwas decanted, the hair rinsedtwice with 0.5 ml water, and the rinsescombinedwith the exposure solutions.Radioactivityin the solution ("unbound isotope")and in hair ("bound

isotope")wascounted.The amountof Zn bound/ghair andthe percentof radioactivity ultimatelyboundto hair wascalculatedfrom the total Zn in the samplesolutions. SUBSTANTIVITY

OF Zn-GLY

TO HAIR:

EFFECTS OF GLYCINE

Two seriesof solutionswerepreparedwith variouslevelsof Zn in both (Figure 3); one serieshad excessglycine present, the other had none. Two ml of each solution was

136

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

allowedto equilibratewith 0.1 g hair for two hours, and I-•mol bound/ghair was calculated.

SUBSTANTIVITY

OF Zn-GLY

TO SKIN

Twopieces ofstratum corneum, 50mgeach,wereexposed to 2 ml of 50mM[65Zn] Zn-GLYwhiletwopieces wereexposed to2 mlof0.01[65Zn]Zn-GLY.Afterdecanting, the stratumcorneumpieceswererinsedtwice with water, and the rinsingswere savedand countedwith the solutionsand stratumcorneumsamples.

INHIBITORY

ACTIVITY

OF RESIDUES

Activity of the enzymesAS and beta-G can be measuredquantitativelyusing the

fluorogenic substrates 4-Me-umbelliferyl sulfate(4-MUS)and4-Me-umbelliferyl glucuronide(4-MUG), respectively, whichare hydrolyzedenzymatically to the fluorescent 4-Me-umbelliferol.Thesematerialsare analoguesof the proposednatural substrates androstenol sulfateand androstenol glucuronide(3). Inhibition of thesereactionsby Zn-GLY boundto hair wasthus measured: A set of 100-mg samplesof hair that had beenexposedto variouslevelsof Zn-GLY was split in half. One half wasplacedin reactionmixtureswith AS (0.001 mg)/4-MUS(0.1 mg) in 2 ml Tris buffer.The other

halfwasplacedin reactionmixtureswith beta-G(0.0001 mg)/4-MUG (0.0025 mg) in 2 ml Tris buffer. Reaction mixtures were irradiated at 334 nm. Fluorescenceemission

intensityat 444 nm wasmonitored overa periodof onehourasdescribed previously (1). COMPATIBILITY

WITH

SOAP PRODUCTS

Solutions of[65Zn]Zn-GLYwereprepared at 0.1 M and0.5 M Zn in water,andalso in solutionsof soapandof syndetbarat variousconcentrations. Samplesof choppedhair (0.1 g each)werethen exposedto 2 ml of eachof thesesolutionsfor varioustime periods, rangingfrom minutesto hours. AXILLARY

DEODORANCY

OF Zn-GLY:

CLINICAL

STUDIES

A completehuman clinical studywasconductedby Hill Top Biolabs,Inc., of Cincinnati, Ohio (7) to assess its efficacy.The testsolutionwas4.5% Zn-GLY (1.5% in Zn); this level representsthe maximum solubility of Zn-GLY in aqueoussolution. The positive control was 5% aluminum chlorhydrate,a level that effectsdeodorancyby antimicrobialactionbut doesnot inhibit perspiration.The negativecontrolwaswater. Odor assessments wereperformedby four experienced professional evaluators from Hill Top Research,using a ten-point scoringsystem.Ninety subjectswere recruited, and

baselinescoreswere established after a twenty-fourhour conditioningperiod. Sixty subjectswith highestscores(all greaterthan 4) were retainedfor the study. Thirty subjectswere assignedto Group I, to test Zn-GLY in one axilla, versuswater in the other; thirty were assignedto Group II, testing aluminum chlorohydratein the same way.

ODOR

INHIBITION

137

During the test the subjectswere enteredinto a regimenof timed, supervised washes with conditioningsoapbars,followedby supervised treatmentswith the test solutions andpost-treatment odorevaluation.The evaluations wereconducted at 5-hr and 24-hr intervalsfollowingtreatments 2, 3, and4, andall aspects wererandomized to eliminate any biases. Scoresof the four judges were averagedfor each person, for each evaluation. The axilla-to-axilladifferencefor eachpersonwastakenasthe indexof the deodorancy of the testmaterialversusthe watercontrol,and thesedifferences weretreatedstatisticallyby the method of Hollander and Wolfe (8).

ANTIMICROBIAL

PROPERTIES

OF ZINC

GLYCINATE

MIC valuesof severalzinc complexeswere measuredfor three organismsthat occur naturallyin the axilla:S. epidermidis, S. aureus,and a lipophilic diphtheroid.

During an in-houseclinicaltrial of Zn-GLY deodorancy, andagainduring the Hill Top Research clinicalstudy,antibacterialpotencywasdeterminedby the methodof Marpies and Kligman (9). 12

Zn GLY at 10 rnM 10

Zn GLY at 1 rnM

Zn GLY 0.01 rnM

0 0

10

20

Time,

min.

Figure 2. Binding of Zn-GLY to hair, kinetics.

3O

138

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

RESULTS PRELIMINARY

STUDY:

KINETICS

OF BINDING

OF Zn-GLY

TO

HAIR

Beforeundertakingquantitativesubstantivitystudies,it wasnecessary to establishthe time scaleof binding of Zn to the substrate.

Figure2 showsthat for Zn concentration exceeding1 mM, bindingis nearequilibrium after an hour and that hair can hold at least 10 Ixmol of Zn-GLY/g. Subsequent experiments wereconductedusingovernightincubationof Zn-GLY with substrate.

SUBSTANTIVITY

OF Zn-GLY

TO

HAIR:

EFFECT

OF GLYCINE

The effectof complexation of the Zn ion on its substantivityto hair wasinvestigatedby measuring the bindingof Zn fromzincchloridein the presence andabsence of glycine. Figure 3 showsthat Zn is boundto hair more efficientlywhen glycineis present. Apparently,Zn-GLY bindsto keratinwith somewhat higheraffinitythanZn chloride.

7O

6O

5O

4O

Zn GLY 3O

Zn Chloride 2O

10

0

1 -4

-3

-2

log Zn GLY concentration,M. Figure 3. Binding of zinc glycinateto hair, concentrationdependence.

ODOR

INHIBITION

Table

139

1

Depositionof Zn-GLY Onto Stratum Corneum

•mol Zn-GLY/g stratumcorneum [Zn-GLY] 0.01

Trial 1

Trial 2

Average

mM

0.3

0.3

0.3

50.0 mM

82.0

108.0

95.0

This studyindicatesthat the formationof Zn-GLY is proceedingin the solutionand that the Zn from Zn-GLY bindsappreciablyto hair from solutionswhereZn-GLY concentration exceeds approximately1 raM. When Zn-GLY expressed as [Zn] exceedsabout 10 mM in solution,binding is about 50 I.•mol/g. SUBSTANTIVITY

OF Zn-GLY

TO SKIN

Table 1 showsthat stratumcorneumexposedto 50 mM Zn-GLY bindsapproximately 90 I.•molZn/g. Exposureto 0.01 mM Zn-GLY produces bindingof lessthan 1 I.•mol/g. 6O

o.oo mM Zn • 50

0.85mMZn

3.4 mM Zn lO

10

20

30

40

50

time, min. Figure 4. Beta-G reactionsin the presenceof hair treatedwith Zn-GLY.

60

7O

140

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

Interestingly,the binding to stratumcorneumis aboutthe sameasthe binding to hair; apparently,the keratinsof botharesimilarwith respectto chemicalfeaturesrecognized during zinc glycinatebinding. INHIBITORY

ACTIVITY

OF RESIDUES

The abovestudiesshowthat exposure to a solutionof 100 mM Zn-GLY is requiredto producea residueof approximately100 }xmol/gof Zn on hair or skin. This concentration of Zn-GLY is probablyattainablein a cosmeticpreparation.We wantedto determinewhetherthe residueleft by suchsolutions wasan effectiveinhibitorof the enzymes responsible for axillaryodorgeneration.

Figure4 shows the timecourses forbeta-Greactions in thepresence of thehairsamples that had beenexposedto variousconcentrations of Zn-GLY. Figure 5 is a similarplot for AS reactions.In both, it is obviousthat the inhibitionof the reactionsis dependent on the concentration of boundZn. Replottedasenzymeactivityvs [Zn] (Figure6), the data showthat hair exposedto Zn-GLY at lessthan 15 mM inhibitsAS by morethan 90%, and that hair exposedto less than 10 mM Zn-GLY inhibits beta-G almost completely.

50

0.00

mM Zn

III

4O

mM Zn'

30

20

10

13.3rnMZn

ß

0

I

10

'

I

20

'

I

30

'

I

40

'

I

50

'

I

60

time, min. Figure 5. Aryl sulfatasereactionsin the presenceof hair treatedwith Zn-GLY.

7O

ODOR

INHIBITION

141

beta-glucuronidase

•

0

I

lO

0

2o

Zn GLY concentration, mM. Figure 6. Inhibition of odor-causing enzymesby hair exposedto variouslevelsof Zn-GLY.

COMPATIBILITY

WITH

SOAP

PRODUCTS

Zn from high concentrationsof Zn-GLY (0.5 M) precipitatesfrom soap solution, probablyas Zn fatty acid salt. Suchinsolubilitysuggeststhat Zn residueson the skin (from a deodorantpreparation)would not be removedby soapor syndetwashing,and would provideenduringinhibition ofß the odor-causing enzymes.We testedthis hy65 pothesisby suspending low concentrations of Zn-labeledZn-GLY in soapand syndet ß 65 ß solutions,exposingfreshhair samplesto the solutions,and measuringthe Zn residue on the hair (seeMethods).Figure7 showsthat dissolved syntheticdetergentbar reduces the binding of Zn-GLY and that soapreducesit evenmore. However,neithersoapnor syndeteliminatesbinding. More than half of the Zn bound from water at any Zn concentration testedis boundfrom test solutionsof soapand syndet.

AXILLARY

DEODORANCY

OF Zn-GLY:

CLINICAL

STUDIES

Basedon thesein vitro studies,which suggestedthat Zn-GLY could be an effective deodorant,a clinical study was conductedby Hill Top Biolabs,Inc., of Cincinnati,

142

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS

syndet,

soap,

•.

0.5 mM Zn% syndet, 0.1 mM Zn

soap, 0.1 mM Zn 0.00

0.05

0.10

0.15

% soapor syndet Figure 7. Reducedbinding of Zn-GLY to hair from soapand syndetcompositions.

Ohio. The studyassessed the deodorantefficacyof Zn-GLY. A test solution1.5% in Zn

(4.5% Zn-GLY) wasemployed; thislevelrepresents themaximumsolubilityof Zn-GLY in aqueoussolution. The positivecontrolwas 5.0% aluminum chlorhydrate.At this levelthe aluminumcompoundeffectsdeodorancy via antimicrobialactionbut doesnot inhibit perspiration.The negativecontrolwasdistilled deionizedwater. A summaryof the resultsof the deodorancy studyis given in Table II. Mean odorscores representthe four-judgeaverageodorscores,post-treatment.The controlvaluesare the averagebaselineodorscoresfor the variouscells. The differences amongtheseevaluationswereanalyzedusingthe distributionfreesigned-ranktestdescribedin reference 8.

Zn-GLY (4.5%) wassignificantlymoreeffectivethan H20 in controllingmalodorat eachpost-treatmentevaluation.Aluminum chlorhydrate(5.0%) wassignificantlymore effectivethan H20 in controllingmalodorat eachpost-treatmentevaluation. The Hill Top report doesnot include a comparisonbetweenthe deodorantefficaciesof Zn-GLY and aluminum chlorhydrate.We thereforeperformedStudent'st-test of the

data (Cricket Statworks).The resultsare given in Table III. The group receiving Zn-GLY treatmentexhibitedlower mean odor scoresthan did the group receiving aluminum chlorhydratetreatment at all six evaluations.Of thesesix differences,four

ODOR

INHIBITION

Table

143

II

Summaryof Resultsof DeodorancyStudy

Application#2

5 hr Zn-GLY

Mean odor score % Difference

3.12

3.13

% Difference

Application#3

H20

ACH

4.75

3.91

H20 5.41

24 hr

2.95

Application#4

4.65

28.1%

H20

Zn-GLY

3.47

3.27

32.7%

% Difference

H20 4.70 30.5%

H20

ACH

4.15

3.88

H20 5.24

28.8%

26.1%

5 hr

24 hr

Zn-GLY 2.60

H20

Zn-GLY

3.94

3.11

34.0%

ACH Mean odor score % Difference

3.34

5 hr

ACH

Mean odor score % Difference

4.03

H20

27.8%

2.33

% Difference

Mean odor score

Zn-GLY

34.0%

Zn-GLY Mean odor score

H20 22.6%

ACH Mean odor score

24 hr

2.90

34.9%

H20 4.74 34.5 %

H20

ACH

4.45

3.64

H20 5.16

29.5%

Baselineodorscores: Zn-GLY, 6.41; H20 , 6.42 ACH, 6.64; H•O, 6.73 Zn-GLY

concentration: 4.5%.

ACH (aluminum chlorhydrate)concentration:5.0%.

weresignificant: all three24-hr evaluations gavesignificantdifferences. Only oneof the 5-hr evaluations gavesignificantdifferences. Zn-GLY appearsto providea strongerand morelong-lastingdeodoranteffectthan doesaluminumchlorhydrateat theselevels. ANTIMICROBIAL

PROPERTIES

OF ZINC

GLYCINATE

Simplesaltsof zinc are well knownto be bacteriostatic. One possibleoutcomeof our work wasdiscovering a wayto reduceaxillaryodorwithoutdestroyingthe naturalflora of the axilla. ThereforeMIC valuesof severalzinc complexesand the antisepticIrgasan DP 300 (positivecontrol)weremeasuredagainstseveralorganismsthat occurnaturally in the underarm, including S. epiderrnid)s, S. aureus,and a lipophilic diphtheroid. Resultsare shownin Table IV. The ">" signsindicatethat bacterialgrowth occurred

144

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Table

III

StatisticalAnalysisof DeodorantEffectof Zn-GLY vs Aluminum Chlorhydrate Mean odor score

Significancelevel

Application2 5 hr

Zn-GLY

3.12

ACH

3.13

0.945

3.34

0.003

24 hr

Zn-GLY ACH

3.91

Application 3 5 hr Zn-GLY

2.33

ACH

2.95

0.005

24 hr Zn-GLY

3.27

ACH

3.88

O.OO3

Application4 5 hr

Zn-GLY

2.60

ACH

2.90

0.155

24 hr Zn-GLY

3.11

ACH

3.64 Table

0.044

I¾

Minimum Inhibitory Concentrations of Zinc CompoundsAgainstVariousBacteria MIC (ppm) Compound

Lipophilic diphtheroid

ZnC12

> 1000

Zn-GLY

> 1000

Zn-NTA

Staphylococcus epidermis > 1000

Staphylococcus aureus > 1000

1000

1000

>9600

>9600

>9600

Zn-HEDPxz

> 1000

> 1000

> 1000

Zn-EDDA

>6500

>6500

>6500

Irgasan

16