J. Soc.Cosmet.
Chem.,29, 537-544 (September1978)
Useof microspectrophotometry
in
dermatological
investigations
GARY L. GROVE, ROBERT M. LAVKER and ALBERT M.
KLIGMAN SimonGreenberg
Foundation,3401 Market Street,
Philadelphia,PA 19104.
Received
January25, 1978. Presented
at AnnualScientific
Meeting,
Society
of Cosmetic
Chemists,December
1977, New York, New York.
Synopsis
Quantitativeestimationof dimensionsof structureand amountsof material in skin at the light microscope
level has,until now, required time-consumingand tediousmethodsthat are often subjectto observererror.
We have overcometheseproblemsby usinga Vickers M86 scanning-integrating
microspectrophotometer.
This analyticallight microscopedetectsthe amount of light which can passthrough a specimenand then
electronicallyconvertsthisvalueinto unitsof absorbance
andprojectedarea.This approachis very versatile
and is in fact applicableto anybiologicalstructurewhichcanbe identifiedat the light microscopelevel andin
which an appropriatechangein color intensitycan be realized.The fundamentalprinciplesof visible light
MICROSPECTROPHOTOMETRY
and its application to DERMATOLOGICAL STUDIES that objectively evaluatethe pathophysiologicalstatusof skin are described.
INTRODUCTION
For many years,light microscopists
have been obliged to rely on "eyeballing"to assess
suchitems as acanthosisor atrophy of the epidermis,enlargementor shrinkageof sebaceousglands,and the amountof dermalgroundsubstance
in histochemicallystained
sections.Awarenessof the crudeness
of suchestimatesled to the developmentof more
objectivemethodssuchasplanimetryand stereographic
grid analysis.Althoughthese
methodsare certainlyimprovements,they alsotend to be tedious,time-consumingand
prone to humanerror. These problemsled us to considermicrospectrophotometryas
an alternativemethod for analyzingdimensionsof structureand amountsof material in
the samples.
Until recently, this technique was considered to be of such an advancednature that
only specializedexpertscould dream of usingit. The advent of commerciallyavailable
instrumentsand improved histochemicalmethods have changedall this. Microspectrophotometryis currently at the heart of severalexcitingbiomedicalresearchprojects
and no doubt the next few yearswill see an increasednumber of instrumentsof this
type beingusedin the routine situationfor screening,diagnosisand other investigatory
purposes.
537
538
JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
24020
.
Figure 1. The Vickers M-85 scanning-integrating
microspectrophotometer,
courtesyof Mr. Robert Osgood, Vickers Instruments,Inc., Woburn, Massachusetts
With this in mind, we would like to describethe fundamentalprinciplesofmicrospectrophotometryand illustratehow a variety of parameters,whichare usefulin assessing
the pathophysiological
statusof humanskin, canthusbe easilyand rapidlymeasured.
Specialemphasiswill be given to the Vickers M-85 scanning-integrating
microspectrophotometer(Figure 1) that we routinely employ for our studies.
INSTRUMENTAL
DESIGN
The principles of microspectrophotometryare similar to those of conventional spectrophotometry(Figure2). Both instrumentsare comprisedof three main units,viz., 1)
DERMATOLOGICAL
INVESTIGATIONS
MACROSPECTROPHOTOM
liable
monochromator
cuvette
light
•
Iolutlon
539
ETRY
photo detection
lyltem
& readout
MICROSPECTROPHOTOMETRY
•
liable
monochromator
light
microscope
•
o
o
photodetection
Syllem & readout
illde
Figure 2. Comparison
of microspectrophotometry
with macrospectrophotometry,
adaptedfrom Chayen
andBitensky(3)
a stablesourceof monochromatic
light, 2) a sampleholderand 3) a photodetection
system.The Beer-LambertLaw,whichdescribesthe exponentialrelationshipbetween
absorptionof monochromatic
light and the amountof absorbingmaterialthe light
traverses, is the basisfor the measurement in both. The differences arise from the na-
ture of the materialbeingmeasured.In conventional
spectrophotometry,
e.g.,Lowry
proteindeterminations,
onemeasures
howmuchlightof a specificwavelengthcanpass
through a cuvette containinga colored solution. In microspectrophotometry,
the
sampleholder is replacedby the opticaltrain of a microscope
allowingmeasurements
to be madeon biologicalspecimens.
In contrastto the homogeneity
of a coloredsolution, the majority of biologicalspecimensare quite heterogeneousand subjectto
markeddistributional
errors.This basicproblemcanbe illustratedby considering
a
squarespecimen(Figure 3) composedof four equal segmentseachof which hasa dif-
ferenttransmittance--an
expression
of howmuchlightcanpassthroughthe specimen.
The relative amount of absorbingmaterial in the specimencan be calculatedas the
productof absorbance
andarea.Note thatby determininganaveragetransmittance
for
the entire specimen,an error of.054 units of 15% has been made. This is the "distribu-
tionalerror" andoccurswhenevera singledirect measurementof intensityismadeon
objectswith regionsof diversetransmittance.
On the otherhand,by calculating
independentlyfor eachregionandsummingthe results,one takesinto accountspecimen
heterogeneityand thusavoidsthe problem of distributionalerrors.
The technicalaspectsof the Vickers M-85 scanningand integratingmicrospectrophotometer have been describedin detail elsewhere (2). In this instrument
(Figure4), the specimenis viewedby a conventional
light microscope
systemandan
adjustablephotoelectricgratingsystemis used to define the field to be measured.
During operationthis field is scannedin a rasterfashionby a flyingspotlightprobe
consisting
of a smallbeamof lightfor whichthe materialexhibitsmaximalabsorption.
540
JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
DISTRIBUTIONAL
ERROR
Equations:
m
where
=
AB
m,, ma•s of absorbing material in specimen
A. absorbance (--log T)
80%
B. area of photometeric
field
T = transmittance
HETEROGENEOUS
OBJECT
MEAN
T
A
.301
INTEGRATED
X
X
B
I
=
=
M
.301 units
T
A
X
B
=
M
!4•.698
x,25
• .175
.398
.222
x
x
.25
.25
.097
x
.25
=
:
•
.100
.056
.024
.355 units
Figure 3. Distributional error in a model system
The specimenis divided into four equal areasof varyingchromophoreconcentrations.The mean
valuewasdeterminedby measuringthe transmittanceof all areassimultaneously
and the integratedvalueby
measuring the transmittance of each area separately. The distributional error increasesas specimen
heterogeneity increases.
At each measuring point the light intensity of the object is transformed into
absorbance by a specially designed analog convertor which makes use of the
logarithmicallydecayingvoltage of a dischargingcondensorto transform the signals
from the photomultiplier into a train of 10 kHz pulses.Since this circuitry simulates
the Beer-Lambert Law, the duration of each of these pulses is proportional to the
absorbance at the point. The digitized value of each sample point within the
electronicallygatedmeasuringfield is storedin a computer.At the end of a scanning
raster involving over 120,000 measurementsof samplearea, eachsmallenoughto be
relatively free of distributional error, these signalsare integrated to give a value proportional to the amount of absorbingmaterial. Simultaneouslya seconddigitalmeter
gives a reading which is proportional to the area of the specimen which has an
absorbancegreater than any arbitrary chosenthreshold value. By usingstandardsof
reference, the absorbance and area meters can be calibrated in absolute units of
picogramsand squaremicrons,respectively.
APPLICATION
USING
ABSORBANCE
MEASUREMENTS
Microspectrophotometrywas originally designedto estimate the DNA content of an
individualcell by measuringthe absorbanceof Feulgen-stainednuclei.This methodhas
DERMATOLOGICAL INVESTIGATIONS
541
INTEGRATED DENSITY & PROJECTED AREA MEASUREMENTS
with ¾1CKERS M85
MICROSPECTROPHOTOMETER
monochromltor
h]
flying-- spot
lister
•
scIn
sc.nner
m.sk
photomultip
SPECIMEN
VIEWING
MASKING
SYSTEM
SYSTEM
specimen
photomultiplier
--
,Integr.ted
Density
signal
cted
Aree
Figure4. Schematic
drawing
showing
relationship
ofthecomponents
ofVickers
M-85scanning-integrating
microspectrophometer
beenextremely
fruitfulinstudies
ofthecellcycle(2).Cellsin G• haveadiploid
or2C
DNA contents,
cellsin G2,a 4C content,
whilecellsin S haveintermediate
values.
Thusthepercentage
ofcellswithDNA contents
exceeding
thediploid
modecanbe
usedto evaluate
the degreeof proliferative
activitysinceit is thesecellsthatare
synthesizing
DNAandpreparing
todivide.
Psoriasis,
ahyperproliferative
skindisease
(4-6),hasbeenstudied
inthismanner.
Asexpected,
therewasamarked
increase
inthe
number
ofhyperdiploid
nucleiin thelesional
skin;ofperhaps
greater
interest
wasthe
finding
thatproliferative
activity
wasalsoelevated
in theclinically
normal-appearing
skinaswell.Weareencouraged
thatthisapproach,
whichobviates
theneedforradioisotopes,
mayprovide
information
thatcanbeofdiagnostic
orprognostic
value.
The Feulgen-DNA
contentdistributions
of tumortissues
oftenreveals
subtle
anomalieswhich aid in the detectionof cancer.Recentlytwo groups(78) have
presented
evidence
whichsuggests
thatmicrospectrophotometry
isuseful
in thecytodiagnosis
of mycosisfungoides,
a malignantskin reticulosis.
Microspectrophotometric
measurements
wereobtained
inthese
studies
fromimprint
specimens
prepared
bytouching
freshbiopsy
material
to a glass
slide.Patients
withclinically
definite
mycosis
fungoides
hadabnormal
Feulgen-DNA
distributions
withaneuploid
andpolyploid
values.
Moreimportantly,
eventhosepatients
in thepremycotic
stage
wholaterwentontodevelop
thisdisease
couldbeprospectively
identified
onthebasis
of subtlebutnonetheless
realdifferences
in theirFeulgen-DNAcontentdistributions.
Thefinalimpact
ofbeing
abletoscreen
formycosis
fungoides
intheearlystages
could
beconsiderable
andwearecurrently
tryingto furtherdevelopthiscytodiagnostic
tool.
542
JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
Table
I
ProceduresAmenableto MicrospectrophotometricMeasurements
1) Histochemical & Cytochemical Staining Reactions:
DNA (Feulgen,Gallocyanin-chromalium,methyl green)
RNA (Azure B, Pyronin Y)
Histones(Alkaline FastGreen, Eosin-FastGreen)
Proteins(Naphthol Yellow S, Millon, Sakaguchi)
Carbohydrates(PAS)
Mucopolysaccharides(Alcian Blue, Mucicarmine, Colloidal iron)
2) EnzymaticHistochemistry:
Lysosomal--BitenskyFragilityTest
Mitochondrial--monoamine
oxidase
PentoseShunt--glucose-6phosphatedehydrogenase
3) Redox State:
Prussianblue of Chevremont-Frederic
4) Natural Pigments:
Cytochrome P~450
Hemoglobin
5) QuantitativeAutoradiography.
Although most absorbanceapplications have centered on Feulgen-DNA measurements, the current surge of interest in microspectrophotometric
analysisprobably
stemsfrom recent developmentin other histochemicalmethods.Although the details
of these proceduresare beyond the scopeof this overview, a few examplesthat are
amenableto microspectrophotometricmeasurementsare listed in Table I. In general,
these methods enable the detection of tissue chemical changes in the picogram
(10-12g) rangewith a routineaccuracyof _+2%.
The use of microspectrophotometryin conjunction with reliable histochemical
methodsoffers severaladvantagesover the conventionalform of biochemicalanalyses
("grind and find"). First, it can be used to make many measurementson minimal
amountsof sampletissue.Moreover, multiparameter analysescan often be achievedin
the same specimen by using a combination of methods either simultaneouslyor
sequentially.The most important advantageoffered by this approachis that it allows
the investigatorto simply relate observedbiochemicalchangesto the structureof the
biologicalspecimenbeing examined.Thus it is quite possibleto measuresuchthingsas
amount of mucopolysaccharides
in the dermis, keratohylin content in the granular
layer, the sudanophiliaof lipids in sebaceous
glands,lysosomalenzymeactivityof the
basallayeror sulfhydrylor disulfidegroupsof keratin in situ.
With the availabilityof suchinstrumentationone no longer needsto be contentwith
making subjective appraisalsof staining intensities. Instead it is now quite easy to
quantify the precise amounts of specificmaterial of a variety of dermatological
specimensfrom normalor diseasedskin.
APPLICATIONS
USING
The Vickers M-85
AREA
MEASUREMENTS
scanning-integrating microspectrophotometer allows one to
measurethe projected area of the specimen.We have found this facility to be
extremelyusefulin histogeometricanalyses.For example,the needfrequentlyarisesto
measurethe mean epidermalthickness,a parameterwhich is markedlyinfluencedby
disease and experimental manipulations. In the conventional approach, the value
DERMATOLOGICAL
INVESTIGATIONS
543
representsthe averageof 25 to 100 readingsat random spotsusingeyepiecegraticule
(10). This is not only time consumingand tedious but is subjectto considerableerrors
in specimenswith prominent fete-ridge patterns.We are not the first to recognizethis
problem, as a method basedon the Quantiment microdensitometerhas been previously reported (11). The key word here is "microdensitometer,"becausethis approach monitors only differencesin grey levels--not color intensities.This presents
problemswhen two or more colorsare present in the samephotometricfield and requires that the sectionsbe overstainedwith hematoxylen,which producesa dark blue
color, enabling the Quantiment to detect these regions from fainter pink dermal
components.With the Vickers M-86 microspectrophotometer
thesecolorscanbe resolvedand projected area measurementsobtainedfor each colored component.Thus,
by measuringthe areaoccupiedby the epidermisin a standardsizefield, one canobtain
a global assessment
of the dimensionsof the viable epidermal compartmentwith little
difficulty.
Muchof ourcurrentresearch
ac.tivity
isconcerned
withdeveloping
noninvasive
testing
proceduresfor monitoringthe physiologicalstatusof skin. One extremelypromising
area is exfoliative cytology,which analyzescells shed from the body surfaces.Ken
McGinley, in our laboratory,hasdeviseda simpledetergent-scrubmethod for quantitative samplingand cytomorphologicalvisualizationof the cellsmakingup the desquamatingportion of the horny layer (12). This approachhasproved to be very valuablein
our studiesof psoriasis(13, 14), aging(15), dandruff (16, 17), contactdermatitis(18)
and steriodatrophy(19). Many of thesestudiesindicatethat changesin corneocytesize
permit a sensitiveevaluation of altered skin physiology,especiallyepidermpoieses.
Unfortunately,sincethesecellstend to be quite irregularin shapethe techniquesthat
have been employed to date to measurethis parameter (axial filar micrometry and
polar planimetry) are subject to considerableerror. We can, however, rapidly and
preciselymeasurechangesin corneocytesize by using the projected area feature of
Vickers microspectrophotometry.
CONCLUSION
The range of applicationsfor both absorbanceand projected area measurements
coveredby this brief surveyhopefully has provided some idea of how valuablethe
microspectrophotometric
approachcan be. By measuringlight absorbancecharacteristics we can analyze dimensionsof structure and amounts of material in any biological
structurethat can be identified at the visiblelight microscopiclevel and in which an appropriate changein color intensitycan be realized. The powerful combinationof fast
and accuarategeometricand absorbancemeasurementsgreatly expandsthe amountof
information which can be obtained from dermatological specimens,viz., scrubs,
biopsies,tissueslices,cultures,etc. In fact, the large number of measurementsavailable and the rapidity with which they can be performed meansthat restraintmust be
exercisedto avoid the unfortunatecircumstanceof being surroundedby reamsof data
which have little relevance to the question at hand.
REFERENCES
(1) R. BarerandF. Smith,Microscopefor weighingbitsof cells,New Sci., 24, 380 (1972).
(2) C. Venderly, Cytophotometryand Histochemistryof the Cell Cycle, in "Cell Cycleand Cancer,"R.
Baserga,Ed., Marcel-Dekker, New York, New York, 1971, pp 227-268.
544
JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
(3) J. Chayenand L. Bitensky,Cell Injury, in "Cell Biologyin Medicine,"E. Bittar,Ed.,JohnWiley and
Sons,New York, New York, 1973, pp 595-678.
(4) W. Wonlrab, Uber den DNS-gehalt von epidermiszellen unbeffalener psoriatikerhaut,Dermatologica,140, 28 (1970).
(5) G. Grove, R. Anderton and J. G. Smith,Jr., Cytophotometricstudiesof epidermalproliferation in
psoriaticand normal skin,J. Invest.Dermatol.,66, 236 ( 1976).
(6) G. Grove and J. G. Smith, Jr., In vivo examinationof cell proliferation in normal and psoriatic epi-
dermisby Feulgen-DNA cytophotometry,in "Psoriasis,
Proceedings
of the SecondInternationalSymposium,"E. M. Farbet and A.J. Cox, Eds.,Yorke, New York, New York, 1977, pp 365-368.
(7) W. Van Vloten, P. van Dujim and A. S. Schaberg,Cytodiagnosticuseof Feulgen-DNA measurements
in cellimprintsfrom skinof patientswith mycosisfungoides,Brit. J. Dermatol.,91,365 (1974).
(8) M. Hagedom and G. Kiefer, DNA content of mycosisfungoidescells,Arch. Derm. Res.,258, 127
(1977).
(9) R. Cowden,A review of applicationfor usersof the Vickers M-85/M-86, scanning
microdensitometer,
VickersTechnicalBulletin, York, England,1976.
(10) D. J. Dykes and R. Marks, Measurementof skin thickness:a comparisonof two in vivo techniqueswith
a conventionalhistometric method,J. Invest.Dermatol.,69, 275 ( 1977).
(11) W. Picton, H. Devitt and M. A. Forgie, Practicalapplicationsof the Quantimet 720, an imageanalyzing
computer,in the field of investigativedermatology,Brit. J. Dermatol.,95,341 (1976).
(12) K.J. McGinley, R. R. MarpiesandG. Piewig,A methodfor visualizingandquantitatingthe desquamat-
ingportion of the humanstratumcomeurn,J. Invest.Dermatol.,53, 107 (1969).
(13) H. W. Goldschmidt, J.J. Leyden andJ. Decherd, Exfoliativecytologyin the diagnosis
of psoriasisof
the nail, Cutis, 10, 107 (1972).
(14) H. W. Goldschmidtand A.M. Kligman, Exfoliativecytologyof humanhorny layer,Arch. Dermatol.,
96, 572 (1967).
(15) G. Piewig, Regionaldifferencesin cell sizesin the humanstratumcomeurnII: effectsof sexand age,
J. Invest.Dermatol.,54, 19 (1970).
(16) J. J. Leyden,K. J. McGinley and A.M. Kligman,Shortermethodsfor evaluatingantidandruffagents,
J. Soc.Cosmet,Chem.,26, 573 (1975).
(17) A.M. Kligman,K. H. McGinley andJ.J. Leyden,The natureof dandruff,J. Soc.Cosmet.
Chem.,27, 111
(1976).
(18) E. Hoelzle and G. Plewig, Effects of dermatitis, stripping and steroidson the morphologyof cor-
neocytes.A new bioassay,J.
Invest.Dermatol.,66, 259 (1976).