extension of the limits of cellular …the intracellular peptidases, particularly leucine...

J. clin. Path. (1958), 11, 520.

EXTENSION OF THE LIMITS OF CELLULARPATHOLOGY: THE ROLE OF ENZYME

HISTOCHEMISTRYBY

A. G. EVERSON PEARSEFrom the Postgraduate Medical School of London

" If we would serve Science we must extend herlimits, not only as far as our own knowledge isconcerned, but in the estimation of others."-R. Virchow (1858)."Cytology cannot be content to go on

indefinitely finding new visible parts of thecell . . . what it needs is to discover methodsthat will describe the organization of cellularevents...."-J. Z. Young (1956).

" New methods must be invented that will besensitive enough to detect early functionalchanges."-G. R. Cameron and S. MuzaffarHasan (1958).My thesis is a simple one, and the three quota-

tions which I have used, a selection from manywhich express the same opinions, illuminate whatI have to say with a commendable economy ofwords.The limits of cellular pathology, drawn up by

Virchow 100 years ago, have been pushed forwardin precisely the way that he suggested, by theapplication of histological techniques. At thattime cellular pathology was not truly cellular buthistological rather than cytological. Few patho-logists are sanguine enough tW foresee furtherextension of pathology by these same histologicalmethods, and it is clear that if advances are to bemade in our knowledge of the mechanisms ofpathological processes then new methods must beinvented which will " describe the organization ofcellular events " and which "will be sensitiveenough to detect early functional changes."

Although Virchow was content to state thatpermanent advances in medicine were made solelyby "discoveries concerning the structure of thebody," the rise of biochemistry, in particular, has-made it clear that analysis of structure without-function is relatively barren. Histochemistry is,concerned particularly in the correlation of struc-.ture with function.We cannot be satisfied, says J. Z. Young, with

identifying the new parts of the cell even by

histochemical methods, and it must be admittedthat information derived from the application ofmethods of analytical histochemistry is in mostcases of limited value. It is to the newer tech-niques of functional histochemistry that we mustlook for help, and it is these alone which cantransform descriptive analytical histopathologyinto a new science of functional cytopathology.This thesis I shall try to develop briefly, and 1shall illustrate it, for the sake of simplicity, solelywith examples from the field of enzyme histo-chemistry. Before doing this, however, it may beexpedient to give details of the present status ofthat youngest division of the relatively old scienceof histochemistry.

If we except the few historical methods forindophenol (cytochrome) oxidase, peroxidase, andthe still current method for DOPA-oxidase, whichwere in existence in the early years of the century,enzyme histochemistry can be considered to datefrom 1939 when George Gomori first describedhis method for alkaline phosphatase. By 1953histochemical methods had been evolved for thedemonstration of 18 different enzymes. Most ofthese belonged to the hydrolytic group and themajority were either carboxylic acid esterases orphosphatases. To-day the figure stands at 45.

This is admittedly not very imposing when con-sidered against the grand total of some 700recorded enzymes (Dixon and Webb, 1958), butthe distribution of these 45 methods among themain enzyme divisions is now much more eventhan it was only a few years ago. This is chieflydue to the development of methods for hydrogentransferring enzymes, particularly the dehydro-genases and diaphorases.

Applied Enzyme HbtochemistryA considerable amount of work has been

carried out by the application of histochemicalenzyme techniques to problems in the field ofcellular pathology. Often the techniques them-

copyright. on June 19, 2020 by guest. P

rotected byhttp://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.11.6.520 on 1 Novem

ber 1958. Dow

nloaded from

http://jcp.bmj.com/

EXTENSION OF THE LIMITS OF CELLULAR PATHOLOGY

selves have been criticized for lack of specificityand sensitivity and for inaccuracy in localization.Many of them clearly need improvement on thisaccount, but until such time as better methods aredeveloped there is much to be said for the employ-ment of existing methods if these can be shown toprovide useful information. Thus, although thedegree of localization provided by the standardtechniques for alkaline phosphatase is less thanone would desire, Novikoff (1955) remarks that itis difficult to see how the distribution of thisenzyme within organs could have been studiedwithout a staining method. Precisely the same

may be said for other enzyme systems, but themapping of these, however attractive and impor-tant it may be to the pure histochemist, has littleappeal to the pathologist who wants to know howand why different processes occur as well as whereand when.There are, nevertheless, a number of indications

for the use of histochemical enzyme techniquesin pathology, and a few examples may be cited toillustrate some of the facets of their use.

General Enzyme HistochemistryIn spite of the fact that a reasonable explana-

tion for the mode of action of the enzyme isseldom forthcoming, the alkaline phosphatasemethod continues to be one of the most popularmethods for application to pathology and to theother basic sciences. A histochemical enzymemethod may be useful, even if the mechanism ofaction of the enzyme is not understood, since itcan act as a sensitive indicator of functionalchanges. Histologists have long sought someindication of the activity of the adrenal cortexother than simple fat staining. Recently, Allenand Slater (1956) have shown that the level ofalkaline phosphatase in the connective tissue andblood vessels of the mouse adrenal'cortex risesafter stimulation of the gland with A.C.T.H.Similarly, Clayton and Hammant (1957) haveobserved an overall diminution of alkalinephosphatasevin the cells of the guinea-pig adrenalcortex after A.C.T.H. In human adrenal corticaltumours those regarded as clinically active are

found to consist of cells having little or no alka-line phosphatase, with an intensely active stromaand blood vessel walls. Thus the results of a

simple enzyme technique may be used to evaluateactivity in respect of corticosteroid production.While a raised level of alkaline phosphatase in

a barrier tissue such as a blood vessel wall usuallysignifies that the transmitting function of thebarrier is increased, no such interpretation can be

put upon changes in acid phosphatase levels.Several other types of activity, however, have beenshown to be accompanied by a rise in the level ofthis enzyme. It has been shown to occur, forinstance, in active phagocytosis (Grogg andPearse, 1952) and in cytolysis (Green and Verney,1956). A fall in the level has been found in mosttypes of prostatic tumour (Woodard, 1952), andthis finding is common enough in respect of manyenzyme systems in many different kinds oftumour for it to have almost the force of a generallaw. That is to say, an enzyme found in largeamounts in the normal parent cell of a tumour islikely to be present in reduced amounts in itsmalignant offspring. The main exceptions to thisrule concern the respiratory enzymes, but apparentexceptions may occur where an enzyme absentfrom the presumed parent cell is present in thecells of the tumour. The probable peptidasefound in carcinoid tumours by Pearse and Pepler(1957) falls into this category.On the whole, acid phosphatase is best used

histochemically as an indicator either of increasedmitochondrial activity or of actual mitochondrialdamage, since a diffuse reaction in properly fixedtissues means that the enzyme has escaped fromits normal environment. This is probably thesignificance of the increase shown by Pearse andMacpherson (1958) in the kidney of the potassium-deficient rat.

Pathological applications of the standardmethods for acetylcholinesterases and pseudo-cholinesterases have been limited in number andrestricted practically to the skin and to the centralnervous system. Glial cells, and particularly thefibrous astrocytes, have been reported to containa strong pseudocholinesterase (Koelle, 1954;Cavanagh, Thompson, and Webster, 1954), butPepler and Pearse (1957) were unable to showthis in human astrocytes or in astrocytomas. Bothof the standard histochemical methods for acetyl-cholinesterases are sensitive, specific, and reliable,and their application to neuropathological prob-lems has already produced useful results.Among the many methods for glycosidases, that

for f3-glucuronidase has been most often appliedto pathological problems. Recent improvementsin one of the standard methods by Fishman andBaker (1956) should encourage its application inthe fields of endocrinology and neoplasia. Onceagain, however, the results may have to be inter-preted in the absence of knowledge as to the exactpart played by the enzyme. According to Monisand Rutenberg (1956), the demonstration of ,B-glucuronidase activity may be useful for the

521



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/

A. G. EVERSON PEARSE

differentiation of epithelial and mesodermaltumours, since the latter do not possess theenzyme.

Diagnostic uses of enzyme histochemistry, asapplied to pathological histology, are at presentnot numerous. This is hardly surprising, however,in view of the short period during which most ofthe applicable methods have been in existence.Further developments can only take place whenthe methods become more or less routine ones inhistopathological laboratories. For this purposethe training of histochemical technicians is essen-tial.Another group of enzymes which has recently

been brought into the histochemical orbit isthe intracellular peptidases, particularly leucineaminopeptidase (Burstone and Folk, 1956;Nachlas, Crawford, and Seligman, 1957). Interesthas centred on the production of the enzyme byinvading tumours (Burstone, 1956) and on therich source of the enzyme found in the stroma ofsuch tumours. Sylven and Malmgren (1955),using biochemical methods, found that the peri-pheral portions of invasive tumours containedhigher cathepsin and dipeptidase activities thanthe older, central areas, and Braun-Falco (1957)has suggested that the production of amino-peptidase may be connected with invasiveness.The activity of leucine aminopeptidase in the

production of parathyroid hormone has been con-sidered by Pearse and Tremblay (1958), who havedemonstrated a presumably direct relationshipbetween the two. The uniform distribution of theenzyme in normal rat parathyroid gland is shownin Fig. 1. The specificity of the histochemicalmethod for leucine aminopeptidase has not beenproved, so that, although it is interesting to specu-late that either the parathyroid hormone, or itscarrier protein, may have a terminal leucyl group,this point cannot be settled by histochemicalmethods. It remains certain that aminopeptidaseactivity is directly related to the production ofparathyroid hormone.The non-specific esterase methods have been

widely used in applied histochemistry, but therehas unfortunately been little effort to distinguishthe various types of esterase activity on the basisof findings and classifications recorded in thebiochemical literature. It is possible, followingthe work of Holt (Holt, 1952; Holt and Withers,1952, 1958), to show with considerable accuracyand reasonably quantitatively the position in thecell of a number of non-specific esterases, usinghalogen-substituted indoxyl acetates as substrates.The evaluation of changes observed in the level

4ll photomicrographs are from 4 or 8 ,u fresh frozen cold microtome(cryostat) sections.

FIG. 1.-Normal rat parathyroid gland (embedded in liver). Showshigh activity of leucine aminopeptidase. L-leucyl-jS-naphthyl-amide. x 120.

and distribution of these enzymes is difficult, how-ever, since their exact function is poorly under-stood. The accurate degree of localization whichthese methods afford, however, and the possibilityof determining small intracellular changes infunction, make it essential to continue with theirapplication on a much wider scale. Now thatthe substrates are more easily available formerobjections to the indoxyl methods must vanish.

It is unfortunately true of the majority ofenzymes for which histochemical methods areavailable that their functional significance isobscure. When we come to the dehydrogenasesand diaphorases, however, we observe a significantdifference. This is because biochemical know-ledge of the function of these enzymes, withinand outside the tricarboxylic acid cycle, is muchmore advanced than it is in the case of the otherenzyme systems. Thus, if we can localize thedehydrogenases histochemically by methods inwhich the amount of colour developed is propor-

522



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


tional to the rate of enzyme activity, we are in aposition, as never before, to draw level with thebiochemist in revealing the mechanisms by whichthe cell carries out its function in health anddisease. We can then say something, at least,about the organization of events within the cell.I propose, therefore, to conclude this paper byconsidering briefly the current histochemical posi-tion in respect of the dehydrogenase systems,since I believe that the most promising develop-ments in modern applied enzyme histochemistryare to be found in this field. Only one other fieldpromises such advances. This is the division ofimmunohistology, based on the original fluorescentantibody methods of Coons and his associates(Coons, Creech, and Jones, 1941; Coons, Creech,Jones, and Berliner, 1942).

Dehydrogenase HistochemistryHistochemical methods for the dehydrogenases

are all derived from the original method ofSeligman and Rutenburg (1951), which wasderived in its turn from the successful use byRutenburg, Gofstein, and Seligman (1950) of theso-called blue tetrazolium salt for the demonstra-tion of dehydrogenase activity in extracts andhomogenates. The principle depends on theacceptance of electrons by the colourless, solubletetrazolium salt and its reduction to a coloured,insoluble, formazan dye.

Rtl-C/\N-R

3 3N N -X

N\

H RN-N-R_ |. H° N- N

Tetrozolium FormazanThe reaction is not reversible in biological

systems, and the amount of formazan depositedcan be related directly to the amount of enzyme

activity (Defendi and Pearson, 1955; Glick andNayyar, 1956). Developments which have takenplace since 1951 have brought the dehydrogenasemethods into the forefront of enzyme histo-chemistry. A series of papers by Farber and hisassociates (Farber, Sternberg, and Dunlap, 1956aand b; Sternberg, Farber, and Dunlap, 1956;Farber and Louviere, 1956) set forth the authors'hypothesis that attempts to localize specificdehydrogenases by tetrazolium methods resultonly in the demonstration of one or other ofthe two pyridine-nucleotide-linked diaphorases.These flavoprotein enzymes transfer hydrogenfrom the substrate to the acceptor, in this case a

3B

dye or tetrazolium salt. The succinic dehydro-genase system forms an exception since it con-tains its own built-in flavoprotein and does notrequire the intervention of a diaphorase intransferring electrons from succinate to tetra-zolium salts.

If Farber's hypothesis is true then the informa-tion which can be derived from the application ofthe tetrazolium methods will be severely limitedin scope. If, on the other hand, the demonstra-tion of specific dehydrogenases is shown to bepossible the whole field is transformed. Thesignificance of such a possibility in applied histo-chemistry can hardly be overestimated.An outstanding drawback of all the tetra-

zolium methods in use up to 1957 was thecrystalline nature of the formazan reaction pro-duct. The most commonly used tetrazolium salts(neotetrazolium, blue tetrazolium) both gave riseto two products, a red one soluble entirely in tissuelipids and a blue or purple one which formed thickneedle-shaped crystals. It was thus impossible toobtain enzyme localization at a cytological level.These objections have recently been overcome, intwo different ways, by the development of newtetrazolium salts.The first of these, nitro-blue tetrazolium (Nitro-

BT), was produced by Tsou, Cheng, Nachlas, andSeligman (1956) and applied by Nachlas, Tsou, deSouza, Cheng, and Seligman (1957) to the intra-cellular localization of succinic dehydrogenase.Nitro-BT is capable of accurate localizationbecause its formazan has a strong binding capacityfor protein (substantivity), and it is thus able'neither to diffuse far from its site of origin nor tocrystallize.

Me N X~N\Me //

NNr

PhCobalt formazan from 3-(4,5-dimethylthiazolyl-2-)2,5-diphenyl

tetrazolium bromide (Beyer and Pyl, 1954).

523



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


An entirely different principle is involved in thesecond method (Pearse, 1957) in which the for-mazan reduction product of a thiazol-substitutedmonotetrazolium salt (Beyer and Pyl, 1954) iscaptured in situ by chelation with a metal ion.The usual metals are cobalt, nickel, and copper,

in that order of preference. Both of the newmethods are capable of localizing enzyme activityat the mitochondrial level (Figs. 2 and 3). In bothcases the deposits of diformazan or cobalt-monoformazan are present in the mitochondriain the form of dots, measuring 0.25 to 0.35 / indiameter (Pearse and Scarpelli, 1958). These aremuch more easily seen with the cobalt-formazantechnique. An average mitochondrion 3 ,u longmay contain four formazan dots; the greatestnumber recorded in a single mitochondrion is six.The view that tetrazolium methods can demon-

strate only the two diaphorases and not the specificdehydrogenases with which they are linked hasbeen shown to be erroneous by Nachlas, Walker,and Seligman (1958) and also by Hess, Scarpelli,and Pearse (1958). It is now possible, in fact, todemonstrate cytochemically the presence of nineof the pyridine-nucleotide-linked dehydrogenases,and these, together with succinic dehydrogenase,are shown in the diagram below by thick arrows.Stages indicated by thin arrows cannot at presentbe shown cytochemically.

Although homogenization and ultracentrifuga-tion techniques have indicated that certain of theoxidative enzymes are intramitochondrial, othersare consistently shown to be present mainly orentirely in other fractions. f3-Hydroxybutyratedehydrogenase, for example, is regarded aswholly mitochondrial, wnile the DPN-linked iso-citrate dehydrogenase appears partly in the mito-chondrial fraction and partly elsewhere. Glucose-6-phosphate dehydrogenase is found exclusivelyin the supernatant. The newer cytochemicalmethods suggest that these three enzymes are allintramitochondrial (see Figs. 2, 4, and 5). Thetwo diaphorases show a similar intramitochondrialpattern of localization (Fig. 6) and the biochemic-ally recorded microsomal TPNH-diaphorase can-not be demonstrated cytochemically by the cobalt-formazan technique (Scarpelli, Hess, and Pearse,1958). Virchow, whose colostrum corpuscle wasa "coherent globule" resulting from the fattydegeneration of an epithelial cell, would perhapshave enjoyed the macrophage from the breast ofa lactating rat, shown in Fig. 4. This contains a' modern" colostrum corpuscle surrounded byshort mitochondria typically rich in glucose-6-phosphate dehydrogenase.The ability to demonstrate effectively so many

of the mechanisms of intracellular oxidation isobviously of great significance in relation to bio-

GLUCOSE

GLYCEROL - o.-GLYCEROPHOSPHATE * TRIOSE PHOSPHATE<- GLUCOSE-6-PHOSPHATE--GLYCOGEN

NEUTRAL LCIATE "* +FAT PYRUVATE PHOSPHOGLUCONIC ACID

ETHANOL ArCETALDEHYDE* 4fFATTY ACID HYDROKY A-KETONIC ACID AACETYL COENZYME A PENTOSEPHOSPHATE CYCLE

ACID OXALOACETATE

+V20,/ CITRATE

MALATE cis- ACONITATE

FUMARATE iso-CITRATE

SUCCINATE az-KETOGLUTARATE- ~ ~~~~++IIOz

CO2 GLUTAMIC ACIDHISTIDINEPROLINEARGIN INE

524



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/

F:....2...;

*: : :.:::: ~~~~~~~~~~~Fn. 3e :

X~~~~~~~~

g~~~~~~~~~~~~ W *FIG. 2 ;

FIG. 3

*::

4

JS~~~~~~

FIG.4~~~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~fG

FIG. 2.-Rat stomach. A single parietal cell. Shows distribution ofDPN-linked isocitrate dehydrogenase activity. Nitro-BT. x 1,840.

Et.3-Rat kidney. Proximal convoluted tubule. Sosscii

.~~~ ~ ~ ~ ~~~ ~FG3.Shw.ucii

dehydrogenase activity in the mitochondria. Cobalt-tetrazolium414method. x 4,300.

FIGta. 4.-Rat breast (lactating). A single macrophage containingW. ~~~~~~~~~(lowerright) a colostrum corpuscle. Shows glucose-6-phosphate

t*. ~~~~dehydrogenase activity in the mitochondria. Cbalt-erziuV method. x 4,300.*1:^ FIG. 5.-Rat kidney. Portions of a number of tubule cells. Shows

i-hydroxybutyrateedehydrogenase activity. At lower right appearsa single mitochondrion containing five formazan dots."Cobalt-tetrazolium method. x 4,300.

FIG. 6.-Rat salivary gland. DPNH-diaphorase activity in aW: single duct cell. Cobalt-tetrazolium method. x 4,300.a. 6 ...

:...FIo. 6

.4i.f

....... -:rV..O'. .!

.4



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


S.

. .

.: .

PLATE IA.-Rat salivary gland. Shows high activity in duct liningcells but poor intracellular localization. Succinic dehydrogenase.Blue totrazolium, mothyl green, x 200.

PLATE Ic.-Rat kidney. Distal convoluted tubules. Similarlyimproved localization provided by an efficient capture reaction.Succinic dehydrogenase. Cobalt-tetrazolium, mothyl green,x 700.

PLATE IB.-Rat salivary gland. Shows great improvement in intra-cellular localization afforded by the use of a tetrazolium salthaving a substantive formazan. Succinic dehydrogenase.Nitro-BT, methyl green, x 200.

-e0

'A.:

At%

PLATE ID.-Mouso cervix. Mothylcholanthrone-induced squamous

cell carcinoma. Growing edge of the tumour. Solochrome

cyanine RS. x 124.

..t 4

.~:W.L.

IM

526

44..

, ;*T.......

9



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


logical and pathological problems. Itbecomes possible to follow small altera-tions in the level of the preferential oxi-dation pathway in single cells and todetect changes from one predominantpathway to another. This cannot beachieved by biochemical methods exceptin the case of homogeneous tissues. Even -then, as Schneider (1956) has emphasized,the result is the average of the contribu-tion of a number of different cell types, 4. 7

and of cells of a single type with widelydiffering enzyme contents. Speaking of <the standard biochemical homogeniza-tion techniques, with which he and othershave made so considerable a contributionto our knowledge of intracellular oxida-tions, Novikoff (1956) says, " It is wellrecognized that the activity of a givenfraction obtained by differential centri-fugation is an average figure." He goesfurther and says, "Even if it were pos- FIGs. 7 asible to free the mitochondrial fraction, a hijfor example, of all contamination by appeother cell particulates, enzymic differ-ences existing among diverse mitochondria with-in a cell, and among different cell types, wouldstill be merged into the average." " This limita-tion," he says, " is particularly troublesome whenthe organism is subjected to altered conditionswhich may shift the relative proportion of differ-ent mitochondria within a cell or of different celltypes in the organ." Dehydrogenase cytochemis-try is free from these limitations though subject tocertain others. Absolute quantitation of enzymeactivity, for instance, cannot at present be achieved.Nevertheless, the sensitivity of the histochemicalmethods is such that ,-hydroxybutyrate activitycan be demonstrated in the particles of Keilin-Hartree preparations although such activity is notdetectable by the usual manometric (Warburg)methods.

Application of the newer histochemical methodsfor dehydrogenases and diaphorases has alreadyshown clearly that the mitochondrial populationis not homogeneous from cell to cell, or evenwithin a single cell, and that this lack of homo-geneity occurs not only with reference to contentof specific enzymes but also with reference tophysical states, such as the permeability of themitochondrial membrane and the sensitivity of theorganelle to environmental influences.We can now show, therefore, not only the

differences between one tissue and another and

FiG. 7 Fi10. 8nd 8.-Part of parathyroid gland of monkey (Macaca irus). Fig. 7 slhowsigh activity of DPNH-diaphorase in a single oxyphil cell. The same cellears in Fig. 8, stained with haematoxylin and eosin. x 900.

between one cell and another, but also betweenone mitochondrion and its neighbour.

Applied Histochemistry of the Dehydrogenasesand Diaphorases

It is unlikely that biochemical studies on theDPN-linked dehydrogenases of the parathyroidgland have yet been carried out. If they have,they will have indicated a moderate level of acti-vity little influenced by physiological changes infunction. Studies on succinic dehydrogenase willshow a similarly moderate and constant level ofactivity. Histochemical studies, on the otherhand (Tremblay and Pearse, 1959), reveal thatDPNH-diaphorase, and five of the histochemicallydemonstrable dehydrogenases which require thisenzyme, are concentrated in the oxyphil cells. Thelatter are not present in the majority of species.In the monkey, however, they are well developedand the glands can be obtained in the fresh statenecessary for the successful application of themethods. Figs. 7 and 8 show a single oxyphilcell, first with its content of intramitochondrialDPNH-diaphorase, and subsequently, after re-moval of the metal formazan, by staining withhaematoxylin and eosin.A similar concentration of DPNH-diaphorase is

shown in Fig. 9, in a symplastic giant cell in asarcoma produced experimentally in a mouse by

3B*

527



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/

A. G. EVERSON PEARSE.. E.: ... >> 5z: W re. a etpt ! . t:., b ....... -- :. ...Mg,>R f iv .... lWt; 2!:> u ., - t i x . t s ' s s.; y; .... . r t .. .. g } 9 x .>rC e e .- .g e sF S* > . A. . , . .e , . . ..

.. f .,iej>,'g ii ; ^.: .j Njg . ;#p :.a ,* .ffi , . ;s 4,. & * s.; B6S! Xe . .* : : :' .Bp.i; . : ; i ':'.: .:. ;s: .

n .;: Nt:o ... A. :S:., ......... . nk, ... ..* .W ' ^ , ., ..'' ^ K ''.. < .S S C > 0. ;._F |>

::.'r.* . .;¢.; e ,ffi, f .. ;, .. *

.# ,E: v nk¢ * %.. x.a,4. ::;: . ..... : z .:rU s il -l b

., -. w i.r >s

s .: jk

FIG. 9.-Mouse tissue. A methylcholanthrene-induced sarcoma.

Shows a high activity of DPNH-diaphorase in a so-called sym-

plastic giant cell. Low activity in the other tumour cells. Cobalt-tetrazolium method, carmalum. x 640.

implan-tation of methylcholanthrene. In bothcases it is tempting to assume that this intenseenzy.me activity is related to increased capacity foroxidation, and that this in its turn is related to thesynthesis of products as yet unknown.

Histochemical studies have also shown thatthere is a special localization of 6-phospho-gluconate dehydrogenase in the macula densaof the normal rat kidney (Nachlas et al.,1958), and this has been confirmed by Hessand Pearse (1959) in the case of this enzymeand also for glucose-6-phosphate dehydrogenase.In experimental hypertension in the rat, inducedby means of a Goldblatt clamp, the clampedkidney shows no activity of glucose-6-phosphatedehydrogenase except in the macula densa and inthe arterial and arteriolar walls. In the un-clamped kidney the amount of enzyme in themacula densa falls to low levels. The appearanceof the normal macula densa in the rat is shown inFig. 10, where the activity of the glucose-6-phosphate dehydrogenase in the mitochondria ofthe five or six cells concerned is evident. If bio-chemical studies were carried out a low level ofenzyme activity would be recorded, and it is pos-sible that in the early stages no difference couldbe shown between the clamped and unclampedkidneys. Hess and Pearse have shown that, as

far as the macula is concerned, the level ofglucose-6-phosphate dehydrogenase rises in theclamped kidney as the development of hyper-tension proceeds. Changes in the rabbit maculadensa in hypertension were recorded by Goor-maghtigh (1939), who described swelling andvacuolation of the cells (marked SI in the

FIG. 10.-Rat kidney. Normal macula densa showsgluco.6-phosphatase activity. Cobalt-tetrazolium method, carmalum.>< 1,000.

diagram below, which is reproduced from hispaper), but no evidence of increased activity has

hitherto been recorded. The juxtaglomerularapparatus (YG) is ultimately involved, in long-standing hypertension, in the general arteriolarincrease in glucose-6-phosphate dehydrogenase.

Further Examples in Applied DehydrogenaseHistochemistry

The inhomogeneity of cells in a single type iswell shown by reference to various types oftumours. Plate ID shows the advancing edge ofa squamous carcinoma of the mouse cervix, pro-hihetobenreode. h jxtgomrua

528



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


FIG. 11

FIG. 12

FIG. 13

t~~~~~~~VK

4

FIG. 11l.-As Plate ID. DPNH-diaphorase reaction. Showswide variation in level of activity in healthy cells. Cobalt-totrazolium method. x 490.

F1G. 13.-AS Fig. 12;. Showrs very- high activity of glutamatedehydrogonase in cells which are about to keratinize.Nitro-BT. x 630.

FIG. 12.-As Fig. 11, but shows a region of keratinization..Low enzyme activity in most cells. Cobalt-tetrazoliummethod. x 475.

FIG. 14.-As Plate ID. Growing edge oftumour. Shows highactivity of glucose-6phosphate dehydrogenase in the.advancing cells. Cobalt-tetrazolium mothod. x 630.

529



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


duced experimentally by implantation of methyl-cholanthrene (Scarpelli and von Haam, 1957),which has been stained in a conventional manner.In the next illustration (Fig. 11) a similar area hasbeen shown to demonstrate the localization ofDPNH-diaphorase in the mitochondria. Histo-logically all the cells present are healthy andviable, and without distinguishing features. Theenzyme method shows, however, that some havelittle activity and others, such as the cell in thelower right-hand part of the picture, are intenselyactive. This difference in activity of DPNH-diaphorase is not unnaturally much greater in thekeratinizing areas of the tumour (Fig. 12) wherethe fully keratinized cells contain little or noenzyme. There is an interesting change obser-vable in cells which are just about to keratinize(Fig. 13), where a pronounced rise in glutamicdehydrogenase is seen. There is no parallel risein any of the other dehydrogenases. An entirelydifferent picture is seen in the growing edges ofthe tumour (Fig. 14). Here, there is a relativelyenormous rise in glucose-6-phosphate dehydro-genase activity and this is closely followed by arise in ribonucleic acid levels within the cells.

Clearly these results are significant and we haveto ask ourselves, or the biochemists, what theymean. In Table I the predominant oxidativemechanism associated with a given enzyme systemis recorded.

TABLE ISUGGESTED SIGNIFICANCE OF DEHYDROGENASE

ACTIVITIES

Mechanism Involved orSubstrate Utilized Co-enzyme Significance

Succinate None Citric acid cycle13-OH butyrate Co I Fatty acid oxidation-+active

acetyl synthesisGlutamate Co I Synthesis or breakdown of

glutamateUrea or creatine formation

Malate Co I Citric acid cycleIsocitrate Co I or f a-Ketoglutarate utilization

Co II i citric acid cycleGlucose-6-phosphate Co II Metabolism via oxidative (hex-

ose monophosphate) shuntsynthesis of f A.T.P.

L Nucleotidesa-Glycerophosphate Co I Glycolysis (early stage)Alcohol Co I GlycolysisLactate Co I Glycolysis (later stage)DPNH Diaphorase Total activity of Co-I-linked

dehydrogenase systemsa-Lipoic acid shunt

TPNH .. Diaphorase Total activity of Co-Il-linkeddehydrogenase systems

Indications of the Physical State of MitochondriaThe mitochondria fresh frozen sections are

altered in some respects by the freezing and thaw-ing involved. They can no longer carry outoxidative phosphorylation, for example. If stepsare taken in histochemical practice to preserve theintegrity of the mitochondria as far as possible, byavoiding heating or drying and by their protectionduring incubation with the use of hypertonicmedia (7.5% polyvinyl pyrrolidone), the appear-ances with the latest tetrazolium methods arethose shown in Figs. 2-6. The mitochondria arestill intact, as can readily be ascertained by inspec-tion with the phase microscope. If for anyreason a mitochondrion is damaged beforeincubation, so that the membrane is abnormallypermeable, the first result is that the reaction rateincreases and more formazan per unit time isdeposited. The appearances of such mitochondriaare recorded in the right-hand tubule of the kidneysection shown in Fig. 15. The middle tubulecontains essentially normal mitochondria. If thedamage is so severe that the mitochondrial mem-branes are ruptured the appearances in thesection approach those seen in the left-handtubule in Fig. 15.

If an osmotic load is put upon the mitochondria,by pre-incubating the fresh frozen sections inhypotonic or otherwise harmful media, thosewhich are abnormally sensitive are observed torespond sooner and more violently than thenormal ones. We thus have a tool which iscapable of showing very early changes in thephysiological state of the mitochondria, and wecan therefore indicate pathological changes in avariety of conditions a long time before anychange can be seen by conventional methods.

In Fig. 16 is shown the TPNH-diaphorasereaction in the juxtamedullary tubules of the renalcortex of a rat maintained for 36 hours on a dietdeficient in magnesium. Swollen (damaged) mito-chondria, sometimes only one per cell, are visible.The renal tubules in other areas show absolutelyno abnormality. The change is more easily seenat higher magnification, as in Fig. 17 (succinicdehydrogenase) and more easily still in Fig. 18,which shows the same tissue pre-incubated for 30minutes in a hypotonic medium. Such treatmentproduces no response from the normal mito-chondria of these tubules (though it may from themore sensitive normal mitochondria of otherorgans such as the parietal cell of the rat stomachshown in Figs. 19 and 20). It is thus possible tocalibrate every type of mitochondrion in everytissue and to apply to it just that load which will

Most of the information in the table is derivedfrom biochemical studies and obviously requiresamplification. Such amplification is beyond thescope of this article.

530



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


FIG. 15.-Rat kidney. Reaction of mitochondria to osmotic injury.This chance observation in a single section shows, left, disruptedmitochondria; right, swollen (damaged) mitochondria containingmore formazan than the normal mitochondria of the centraltubule. Succinic dehydrogenase. Cobalt-tetrazolium method.x 850.

FIG. 16.-Rat kidney. Juxtamedullary region of the cortex. Threedays Mg deficiency. Shows early change in mitochondria(swelling and increased reaction rate). TPNH-diaphorase.Cobalt-tetrazolium method. x 750.

Fla. 15

A

*4.4 a

Fia. 16..

>4<4:.x

.,iE' :

...*4$,

....,

j; f

X.

*J. .. .f

FIG. 18

FIG. 17.-Rat kidney. Succinic dehydrogenase reaction afterthree days Mg deficiency. Shows early damage to mito-chondria in tubule cells of the juxtamedullary region.Cobalt-tetrazolium method. x 900.

FIG. 18.-As Fig. 17, but section pre-incubated in hypotonicmedium before succinic dehydrogenase reaction.Abnormal mitochondria show larger, darker formazandeposits. x 900.

Fla. 17

531

V

f *,.

Vol -

... 00,4



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


FIG. 19

S 9.,

6.M

*N

-;S

9.t we

:9

a..

.^0.A..FS.

:4.4*

FIG. 21

FIG. 19.-Rat stomach. Parietal cell incubated for succinicdehydrogenase in a protective medium. Shows mito-chondrial localization of enzyme. x 4,450.

FIG. 21.-Rat kidney. Early stage of Mg deficiency (fifthday). Shows fatty degeneration. Oil red 0. x 900.

FIG. 22

FIG. 20.-As Fig. 19, but incubated in hypotonic medium.Individual mitochondria are entirely unrecognizable.

4,450.

FIG. 22.-Rat kidney. Similar stage to Fig. 21. Shows closeassociation of early stages of fatty degeneration (clearhalos) with mitochondrial damage or swelling. DPNH-diaphorase. x 900.

FiG. 20

I,

A

532

... }

... * M.



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


reveal the earliest change in sensitivity (per-meability) of the membrane. The implications ofsuch a method in experimental pathology needlittle amplification.The gross changes in the kidney of magnesium

deficiency, which include intracellular and inter-stitial calcification, become visible by standardtechniques after a period of about nine days.They occur in precisely the area delineated by"' mitochondrial assay " a matter of hoursafter the start of the experiment. Biochemicalchanges are noted on the fifth or sixth day, whenthe serum magnesium figures fall to low values.The " blindness " of the usual histological tech-niques needs no further emphasis.A similarly early appreciation of the patho-

logical changes has been observed in every con-dition to which the methods of " mitochondrialassay" have been applied. In poisoning withdiphtheria toxin, for example, changes are visiblein the sarcosomes of the cardiac muscle twohours after injection of the toxin. This compareswith the seven days of the usual techniques. Inrenal calcification, induced by single doses ofvitamin D, changes are similarly visible in a fewhours in the region of the kidney where ultimately(nine to 15 days) massive calcification occurs. It ispossible, incidentally, using sensitive fluorescenttechniques for calcium (Pearse, 1959), to detect thedeposition of calcium in the interstitial tissues andin some of the cells of the renal tubules by thefifth day of the experiment. Starvation is anothercondition which induces changes in animal tissuesand these can likewise be appreciated by " mito-chondrial assay" after only a few hours' depriva-tion of food (in the case of the rat). It isinteresting to note that the fatty degeneration,described and studied by Dible and Gerrard(1938), and by other workers (Fig. 21), appears tobe closely connected with the swelling of damagedmitochondria. This is seen in Fig. 22, wherethe swollen mitochondria which appear as fatdroplets in the preparation stained with oil redare seen to contain small circular regions offormazan deposition. This is not the appearanceof normal fat droplets, which are at firstcolourless, though they may subsequently developautoxidation products which will reduce adsorbedtetrazolium salts if these are lipid soluble.

ConclusionEnzyme histochemistry offers many things, first

and foremost, an escape from misleading concep-tions due to consideration of tissues as homo-geneous. Dixon and Webb (1958), in their

masterly work on " Enzymes," say that " it is byno means certain that even cells which are histo-logically of the same type have the same enzymiccomposition." In another passage they maintainthat " unfortunately no information on theenzymic homogeneity of tissues is available."Enzyme histochemistry, on the other hand, demon-strates with certainty that histologically identicalcells frequently have entirely different enzymiccomposition (examples have been given in thispaper) and one can say that if " information onthe enzymic homogeneity of tissues" is lackingthere is already a considerable volume of informa-tion on their enzymic lack of homogeneity.The diaphorase and dehydrogenase methods, a

combination of DPNH-diaphorase and succinicdehydrogenase for example, provide a simple andreliable method for demonstrating the mito-chondrial population of mammalian, invertebrate,and plant cells. Nobody should further considerfor an instant the employment of older stainingtechniques for these organelles.Although the true pattern of metabolic activity

in cells will only become apparent in most caseswhen much more exploratory work has beendone, one can say, with little exaggeration, that ifa single mitochondrion in an organ, or at least ina section from that organ, is abnormal one candemonstrate this fact and determine to someextent the metabolic direction of this abnormality.Such is the sensitivity of the techniques which

I have described that there can be no doubt thatprobably more than 50% of all the studies inexpzrimental pathology, carried out by whatevermethods since the time of Virchow, can berepeated to-day with advantage and with theabsolute certainty that new understanding of " theorganization of cellular events" will result. Afurther application of these techniques can beforeseen in the field of experimental pharma-cology where it will be possible to demonstrate theearliest effects of drugs upon cells in sections,smears, or tissue cultures, with obvious economyof effort.

I could wish that Rudolf Virchow were hereto-day to describe for himself the possibilities forthe extension of his new science of cellularpathology which I have tried, inadequately, to setforth in this article.

It should be evident, from the bibliography, thatthe results reported in this article are not "all myown work." I should nevertheless like to acknow-ledge the great help I have received from my col-leagues, particularly Drs. Defendi, Grogg, Guerrero,Hess, Macpherson, Pepler, Scarpelli, and Tremblay.

533



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/


The photomicrographs are the work of Mr. W.Brackenbury, and the drawings were done by Mr.F. G. Saunders.

REFERENCESAllen, J. M., and Slater, J. J. (1956). J. Histochemn. Cytochenm., 4, 1 10.Beyer, H., and Pyl, T. (1954). Chem. Ber., 87, 1505.Braun-Falco, 0. (1957). Klin. Wschr., 35, 50.Burstone, M. S. (1956). J. nat. Cancer Inst., 16, 1149.

and Folk, J. E. (1956). J. Histochem. Cytochem., 4, 217.Cameron, G. R., and Muzaffar Hasan, S. (1958). J. Path. Bact., 75,

333.Cavanagh, J. B., Thompson, R. H. S., and Webster, G. R. (1954).

Quart. J. exp. Physiol., 39, 185.Clayton, B. E., and Hammant, J. E. (1957). J. Endocr., 15, 1.Coons, A. H., Creech, H. J., and Jones, R. B. (1941). Pr-oc. Soc. exp.

Biol. (N.Y.), 47, 200.and Berliner, E. (1942). J. Immunol., 45, 159.

Defendi, V., and Pearson, B. (1955). J. Histochem. Cytochem., 3, 61.Dible, J. H., and Gerrard, W. W. (1938). J. Path. Bact., 46, 77.Dixon, M., and Webb, E. C. (1958). Enzvmes. Longmans, London.Farber, E., and Louviere, C. D. (1956). J. Histochem. Cytochenz.,

4, 347.Sternberg, W. H., and Dunlap, C. E. (1956a). Ibid., 4, 254.

(1956b). Ibid., 4, 284.Fishman, W. H., and Baker, J. R. (1956). Ibid., 4, 570.Glick, D., and Nayyar, S. M. (1956). Ibid., 4, 389.Gomori, G. (1939). Proc. Soc. exp. Biol. (N. Y.), 42, 23.Goormaghtigh, N. (1939). Brux-med., 19, 1541.Green, M. H., and Verney, E. L. (1956). J. Histochem. Cytochenm.,

4, 106.Grogg, E., and Pearse, A. G. E. (1952). Brit. J. exp. Path., 33, 567.Hess, R., and Pearse, A. G. E. (1959). J. exp. Med. In the press.

Scarpelli, D. G., and Pearse, A. G. E. (1958). Nature (Lond.),181, 1531.

Holt, S. J. (1952). Ibid., 169, 271.and Withers, R. F. J. (1952). Ibid., 170, 1012.-(1958). Proc. roy. Soc. B., 148, 520.

Koelle, G. B. (1954). J. comp. Neurol., 100, 211.Monis, B., and Rutenberg, A. M. (1956). J. Histochem. Cytocheni.,

4, 498.Nachlas, M. M.. Crawford, D. T., and Seligman, A. M. (1957).

Ibid., 5, 264.Tsou, K.-C., de Souza, E., Cheng, C.-S., and Seligrran, A. M.

(1957). Ibid., 5, 420.Walker, D. G., and Seligman, A. M. (1958). J. biophys. biochem.Cytol., 4, 29.

Novikoff, A. B. (1955). In Analytical Cytology, ed. R. C. Mellors,McGraw-Hill, New York.

--(1956). Proc. 3rd int. Congr. Biochemistry, Brussels. 1955,p. 315. Academic Press, New York.

Pearse, A. G. E. (1957). J. Histo(hem. Cytochemn., 5, 515.(1959). Histocheinistry, Theoretical and Applied, 2nd ed. J. & A.Churchill, London. In the press.

-and Macpherson, C. R. (1958). J. Path. Bact., 75, 69.and Pepler, W. J. (1957). Nature (Lond.), 179, 589.and Scarpelli, D. G. (1958). Ibid., 181, 702.and Tremblay, G. (1958). Ibid., 181, 1532.

Pepler, W. J., and Pearse, A. G. E. (1957). J. Neurochem., 1, 193.Ru;enburg, A. M., Gofstein, R., and Seligman, A. M. (1950). Cancer

Res., 10, 113.Scarpelli, D. G., and von Haam, E. (1957). Amer. J. Path., 33, 1059.

Hess, R., and Pearse, A. G. E. (1958). J. biophys. biochenm.Cytol., 4.

Schneider, W. C. (1956). Proc. 3rd int. Congr. Biochemistry. Brussels,1955, p. 305. Academic Press, New York.

Seligman, A. M.. and Rutenburg, A. M. (1951). Science, 113, 317.Sternberg, W. H., Farber, E., and Dunlap, C. F. (1956). J. Histochem.

Cytochem., 4, 266.Sylven, B., and Malmgren. H. (1955). Exp. Cell Re.., 8, 575.Tremblay, G., and Pearse, A. G. E. (1959). Brit. J. exp. Path. In

the press.Tsou, K.-C., Cheng, C.-S., Nachlas, M. M., and Seligman. A. M.

(1956). J. Amer. chem. Soc., 78, 6139.Virchow, R. (1858). Cellular Pathology (English translation by F.

Chance, 1860). Churchill, London.Woodard, H. Q. (1952). Cancer, 5, 236.Young, J. Z. (1956). Endeavour, 15, 5.

534



j.com/

J Clin P


ber 1958. Dow

nloaded from

http://jcp.bmj.com/

extension of the limits of cellular …the intracellular peptidases, particularly leucine...

Documents