J Clin Pathol 1987;40:1194-1200

Glycol methacrylate embedding for light microscopy Ienzyme histochemistry on semithin sections ofundecalcified marrow cores

A ISLAM, E S HENDERSON

From the Department ofMedical Oncology, Roswell Park Memorial Institute, Buffalo, New York, UnitedStates ofAmerica

SUMMARY A simple, routine procedure for water miscible glycol methacrylate (GMA) embeddingof undecalcified bone marrow cores, which preserves the activity of enzymes useful in diagnosingvarious haematopoietic disorders, is described. The GMA used in this study has a low acid contentthat eliminates background staining, and the modified May-Grunwald-Giemsa stain provides gooddefinition and excellent colour differentiation of various haematopoietic cells in the bone marrow,thereby providing optimal conditions for the study of the morphology and enzyme activity of bonemarrow cells in the same preparation. The method is simple, reproducible, requires no expensiveequipment, and is suitable for routine processing of small bone marrow cores in any histopathologyor haematology laboratory.

Enzyme markers are useful in characterising leu-kaemic cells, particularly in the acute types.1 2 Theuse of enzyme markers to diagnose haematologicalmalignancies, however, has been restricted to bonemarrow aspirates or to touch preparations. Cyto-chemical reactions performed on bone marrowbiopsy specimens could be very informative, becausein histological preparations tissue organisation ismaintained and the sampling errors inherent in bonemarrow aspiration are avoided. These methods couldbe particularly important in analysing focal lesions ofbone marrow, or when aspirated material isunobtainable.

Unfortunately, the superior histological techniqueof methyl-methacrylate embedding3 has beenunsuccessful when used with enzyme histochemistrymethods. Water miscible glycol-methacrylate (GMA)embedding, however, which was introduced by Rud-dell,6 provides good morphological detail and yieldsexcellent histochemical results.7'8 It also has goodstretching and mounting properties and, whenapplied to small size marrow cores, can provide sec-tions with excellent preservation of a wide variety ofenzymes used to diagnose and classify haematologicalmalignancies.9 10The value of GMA embedding in diagnosing hae-

Accepted for publication 9 April 1987

mopoietic disorders has been recognised for someyears; its use, however, has been restricted, becausethe technique entails procedures that are not practicalfor histology laboratories in most general hospitals.

This paper describes a simple GMA embeddingtechnique that requires no expensive microtome orelaborate apparatus, and can be easily adapted forroutine use in a histopathology or haematology labo-ratory. It provides more precise histochemical local-isation of enzyme activity than frozen or paraffinsections. In combination with the May-Grunwald-Giemsa (MGG) stain, haematopoietic tissue cells canbe differentiated almost as well as in Romanowsky'sstained smears of aspirated marrow. The MGG stain-ing procedure is more rapid and less complicated thanthe usual Giemsa or haematoxylin and eosin pro-cedures, though the results are similar to the routinelyused Romanowsky's stain. Moreover, the stainingprocedure does not stain the GMA embeddingmaterial, thus avoiding any background colour.

Material and methods

Pieces of bone marrow cores (8-10 mm) wereobtained from patients using a Jamshidi' or Islam"2needle, or one of the disposable needles, and werefixed immediately in PGA (1% paraformaldehyde,0 25% glutaraldehyde, and 0 25% acrolein in 0 1 M

1194

copyright. on A

pril 21, 2021 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.40.10.1194 on 1 O

ctober 1987. Dow

nloaded from

Glycol methacrylate embedding for light microscopy

sodium cacodylate buffer at pH 7-4)9 for one hour at4°C on a rotor mixer. Long (20-22mm or over) mar-row cores were not only difficult to process in GMAbut were also difficult to section. Thus when longmarrow cores were obtained they were divided intotwo smaller pieces and then fixed. After fixation mar-row cores were washed for one hour at 4°C in 0-1 Msodium cacodylate buffer at pH 7-4 with 3% sucrose.They were then dehydrated at 4°C for 10 minutes in50% acetone with 50% 0 1 M sodium cacodylatebuffer containing 3% sucrose, followed by 10 minutesin 75% acetone and 10 minutes in 100% acetone.The dehydrated tissues were placed in solution A

(hydroxyethyl-methacrylate 100ml, 2-butoxyethanol15 ml, and benzoyl peroxide -0g) at 40C for twohours under vacuum. After two hours solution A wasrenewed and the specimens were left in fresh solutionA for 24 hours under vacuum. After this infiltrationthe specimens were transferred to completeembedding mixture-42 parts solution A and onepart solution B (polyethylene glycol 400, 10 ml, andNN dimethylanilin 1 ml)-placed in embeddingmoulds (Sorvall), and left to polymerise overnight at40C under vacuum. The polymerisation reaction isstrongly exothermic and if allowed to proceed rapidlywithout adequate cooling, the boiling point of themonomer may be exceeded with resulting formationof gas bubbles which can be trapped in the plastic andseriously impair the properties of the block. Thisexcess rise of temperature not only damages the tissueform overheating but also destroys the activity of theenzymes. The presence of even a single gas bubble inthe block must be taken as a sign that temperatureshave exceeded the boiling point of the monomer. Thepolymerisation reaction is also inhibited by low tem-peratures and oxygen. Thus polymerisation at 4°Cunder vacuum provides an effective cooling systemand permits slow but uniform and adequate poly-merisation.

Semithin sections were then cut from theundecalcified bone marrow cores with a Tungsten-carbide knife in a Jung's high performance microtome(Autocut), floated on water, placed on glass slides,and allowed to air dry at room temperature. The sec-tions were stained for myeloperoxidase activity using3.3' diaminobenzidine tetra hydrochloride (DAB) asthe substrate'3 and counterstained with Cole's hae-matoxylin. The slides were then washed in distilledwater, allowed to dry, and mounted with diatex. Thesections were also stained for acid phosphatase by theGoldberg and Barka method"4 and for acidoa-naphthyl acetate esterase by the method of Yamet al. l5

For MGG staining, the sections were placed inMay-Grunwald solution (MG stain 2 ml + 48 mlSorenson's phosphate buffer pH 6-8) and incubated in

1195

a dry oven at 45°C for 45 minutes. The sections werethen transferred directly into Giemsa solution(Giemsa stain I ml + 49 ml of Sorenson's phosphatebuffer pH 6 8) and incubated at 45°C for 45 minutes.The slides were washed briefly in distilled water, gen-tly blotted dry, cleared in xylene, and mounted withdiatex.The MG and G solutions and the mounting

medium diatex were obtained from Raymond ALamb, London, and the Sorenson's phosphate bufferfrom Fisher Scientific, Springfield, New Jersey. Inaddition to MGG staining, most other conventionalstains used on paraffin embedded tissue section can beused with good results.

Results

This method of infiltration and GMA embedding ofbone marrow cores is relatively simple, does not con-tain holes or air bubbles (fig 1), is easy to section, anddoes not require longer time than standard methods.This technique provides consistently good quality his-tological sections of 2-3 gm in thickness fromundecalcified bone marrow cores. One micrometrethick sections were also possible, but they were toothin to handle and did not bind sufficient dye for easyinterpretation.

Using GMA, the methods of Leduce andBernhard'6 and Spaur and Moriarty17 were tried.

Fig I Surface view ofbone marrow biopsy core embedded inGMA. Note centrally placed marrow core and absence ofanyholes or air bubbles in this plastic block.

copyright. on A

pril 21, 2021 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.40.10.1194 on 1 O

ctober 1987. Dow

nloaded from

Islam, Henderson

^-F islw w -w__ w- -

M41'~~~~~~~4 1 * _^.

:6 __

i 00~~~~~* a S *^Fig 2 Detail ofbone marrow biopsy preparationfrom patient with non-Hodgkin's lymphoma: (a) after decalcification andparaffin embedding; (b) after embedding in GMA without decalcification. Note cellular detail and excellent tissue preservationin this preparation. Plasma cells (arrows) can be easily recognised. (a) Haetoxylin and eosin; (b) May-Grinwald-Giemsastain.

These authors advocated prepolymer formation fromthe final embedding mixture to reduce the swellingartefacts in the tissue during polymerisation, but con-sistent uniform prepolymer formation by their tech-niques was difficult. The heat released duringprepolymer formation often drove the mixture intospontaneous polymerisation, and the Leduce andBernhard'6 technique of prepolymer formation overa Bunsen burner in a capped Erlenmeyer flask wastoo dangerous. The methods of Sims'8 and Lee'9were much more satisfactory, but this resin failed to

polymerise uniformly, even after air was excluded andthe temperature raised. Attempts to apply the tech-nique to larger moulds to process the whole bonemarrow cores were unsuccessful. Beckstead andBainton9 and Westen et al20 have combined methyl-methacrylate with GMA to produce an embeddingmedium suitable for enzyme histochemistry, but it isnot clear why both media are required when similaror better results can be obtained with GMAembedding alone; and our experience with the tech-nique of Beckstead and Bainton for embedding and

1196

copyright. on A

pril 21, 2021 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.40.10.1194 on 1 O

ctober 1987. Dow

nloaded from

Glycol methacrylate embedding for light microscopysectioning of undecalcified marrow cores has been lesssatisfactory. According to our tests the GMAembedding technique described by Germain was themost satisfactory.2 ' Bone shattering was minimal andthe cytological details of haemopoietic tissue werewell preserved. The consistency of the final block,however, was too hard and brittle for adequate sec-tioning in the Autocut microtome.

Consequently, various combinations of the com-ponents of embedding medium were tested22 and,after repeated trials, a mixture was achieved that pro-duces blocks free of air bubbles and was of a consis-tency that could be cut at any thickness from 30 gm toI gm with a Tungsten-carbide knife in a ReichertJung Microtome (Autocut). Sections as thick as20-30m were necessary to trim the block to reach anadequate depth before the final sections of 2-3 pimwere cut.

Early in this study several fixatives-para-formaldehyde, gluteraldehyde, phosphate bufferedformalin, Schaffer's solution,23 Bouin's solution andPGA-were tested22 as fixative for tissue embeddedin GMA to find the best fixative for routine andspecial stains for enzyme histochemistry. TheSchaffer's solution produced the best nuclear andcytoplasmic detail which facilitated recognition of celltypes in haematopoietic tissue in the sections of bonemarrow cores stained by MGG, but tissues fixed inPGA or Bouin's solution at 4°C for one hour were themost satisfactory for enzyme histochemistry.

Unlike many other plastics, GMA cannot be dis-solved out of the section by any technique available atthe present time, and the unremoved resin stainsheavily with most basic dyes, thus obscuring thesharpness of cellular details. It has been suggestedthat this background staining can be removed com-pletely from the matrix by differentiating the sectionin either 2-butoxyethanol or absolute ethyl alcohol,or in mixtures of these two solvents.6 In our hands,however, such treatment often produced unevenstaining and caused detachment of sections from theslides. Among the various GMA's tested,22 the oneobtained from Hartung Associates was the most satis-factory, and its low acid content produced little or nobackground staining. Thus our combination of Har-tung's GMA and the modified MGG staining tech-nique described herein produced the bestRomanowsky's staining for sections of undecalcifiedmarrow cores embedded in GMA.

Fig 2 compares the details of bone marrow biopsysections from a patient with non-Hodgkin's lym-phoma after GMA and conventional paraffinembedding techniques. One half of the biopsy speci-men was fixed in formaldehyde solution, decalcified,and processed into paraffin (fig 2a); the other half wasfixed in PGA and processed into GMA without

1197decalcification (fig 2b). The superiority ofGMA overparaffin in preserving morphological detail is evident.The plasmal cell infiltration, which was missed in thehaematoxylin and eosin stained paraffin-embeddedsections, was clearly shown in sections stained withMGG and embedded in GMA. The MGG stain pro-vided good definition and excellent colourdifferentiation of various haematopoietic cells foreasy and accurate identification.The enzyme myeloperoxidase (MPO) is a good

marker for the granulocytic series. By light micro-scopy, the Hanker method, using 3-3'diaminobenzidine (DAB) at pH 7-6, seemed to bemost sensitive.'3 It was not difficult to showmyeloperoxidase activity in GMA embedded plasticsections. The reaction product was shown in all thecells of granulopoietic series, including myeloblasts(fig 3) and mature polymorphonuclear neutrophils,after only 10 minutes of incubation. The lympho-blasts were negative; the enzyme was also very activein mast cells, basophilic granulocytes, andeosinophilic granulocytes. Peroxidase rich, largergranules of these cells exhibited more intense brown-ish coloration (fig 5). Enzyme activity was also shownin cells of monocytic lineage, but the reaction prod-ucts were more finely granular, with less intense col-oration.Enzyme acid phosphatase activity was also easily

shown in the sections of GMA embedded bone mar-row cores, and the reaction product was especiallyheavy in osteoclasts (fig 6) and macrophages. Mono-cytes showed considerable variation in staining reac-tions, and megakaryocytes were mainly unreactive.Tartrate resistant acid phosphatase was also shown inGMA embedded sections of marrow cores frompatients with hairy cell leukaemia (fig 7). Enzyme acidalphanaphthyl acetate esterase (ANAE) activitymainly followed the distribution pattern of acid phos-phatase. The reaction product was particularly heavyin monocytes and macrophages (fig 8). A diffuse pos-itive reaction was also observed in megakaryocyticcells, but the reaction product was much weaker.

Discussion

Improved plastic embedding techniques that producereliable, semithin sections of high histological qualityfrom undecalcified bone marrow cores for lightmicroscopic study provide for better evaluation of themorphological and structural detail of the bone mar-row in various haematopoietic disorders. For lightmicroscopic examination of bone and bone marrowspecimens, embedding in methyl-methacrylate is themethod of choice, as decalcification is avoided andsemithin sectioning of large specimens is possible.5Unfortunately, this technique is not successful for

copyright. on A

pril 21, 2021 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.40.10.1194 on 1 O

ctober 1987. Dow

nloaded from

Islam, Henderson

;r)

.S

S~~ ~ ~ ~

V. .., d

16 a,

I .

44

., t. A

4.. '

0

v i

:.. .mb

t6

t

O.isxx

O;*

Fig 3 Bone marrow biopsy sectionfrom case ofacutemyeloidleukaemia showing myeloperoxidase positive blastcells. Bone marrow specimen was processed into GMA andstainedfor MPO by Hanker's method.

9, 4

I 2:if .I4 **0'....t%,:.

-.*....!4*w a

.

Fig 4 Bone marrow biopsy sectionfrom a case ofCML inlymphoblastic transformation shows afocus myeloperoxidasenegative blast cell. Note peroxidase positive maturegranulocytic precursor cells at periphery of lesion. Thisspecimen was processed into GMA and stainedformyeloperoxidase by Hanker's method.

I' t1,.

tV

:*

4s.

Fig 5 Bone marrow biopsy sectionfrom case ofchronicmyeloid leukaemia in chronic phase showing intenseperoxidase activity in eosinophilic granulocytes. Biopsyspecimen was processed into GMA and stainedformyeloperoxidase by Hanker's method.

4..4iOpr

.,m

Fig 6 Biopsy section reactedfor enzyme acidphosphatase(AP). Note strong positive reaction in osteoclasts. BT =

bony trabecula.

,-.w !-.ft:b1

0-0 :,I

4

Fig 7 Bone marrow biopsy specimenfrom newly diagnosedcase ofhairy cell leukaemia showing tartrate resistant acidphosphatase positive "hairy" cells. Bone marrow biopsyspecimen was processed into GMA and stainedfor AP inpresence of tartaric acid.

Fig 8 Bone marrow biopsy section reactedfor enzyme acida-naphthyl acetate esterase (ANAE). Note strong ANAEreaction in macrophages.

1198

.11.

--k"0

copyright. on A

pril 21, 2021 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.40.10.1194 on 1 O

ctober 1987. Dow

nloaded from

Glycol methacrylate embedding for light microscopy 1199

enzyme histochemistry. GMA embedding is the mosteffective medium for preserving optimal mor-phological details and staining enzymes.

This cold GMA embedding technique for pro-cessing undecalcified marrow cores shows the pres-ence of several enzymes useful in diagnosing varioushaematopoietic disorders using special stains and his-tochemical techniques; and the new MGG stainingproduces the best Romanowsky's staining for sec-tions of undecalcified marrow cores embedded inGMA. Thus the morphology and enzymatic com-position ofmarrow cells in their own natural environ-ment can be studied, and on the sections from thesame plastic embedded bone marrow biopsy speci-men. In this way valuable information, which mightotherwise go unnoticed, can be obtained.

Myeloperoxidase (MPO) is an enzyme foundprimarily in the cells of granulocytic series and hasnever been reported in lymphoid cells. This featuremakes it useful for distinguishing between lymphoidand non-lymphoid leukaemias. This reaction, whenperformed on sections of bone marrow biopsy speci-mens, gives accurate information on the nature, local-isation, homogeneity, and extent of leukaemicinfiltrate in the marrow. With this enzyme marked onsections, it is possible to identify more precisely thespatial localisation of granuloid precursors in relationto bone, vascular, and fatty stroma at diagnosis andduring haemopoietic regeneration following chemo-therapy for acute myeloid leukaemia, which may pro-vide some insight into the origin and migration ofcells of granulocytic series within the marrow cav-ity.24 25 These findings may have important biologi-cal implications.The enzyme acid phosphatase is a valuable marker

for lymphoid malignancies and helps distinguishdifferent types of acute and chronic lympho-proliferative disorders. It is important in character-ising T and non-T acute lymphoblastic leukaemia.26The acid phosphatase reaction is positive in blast cellsof most (about 90%)27 of T derived acute lympho-blastic leukaemia with typical localisation in a smallperinuclear area. In contrast, in most non-T-ALL(about 90%), this reaction is absent or weak, andpresent in only a small proportion of blast cells.27 Inchronic lymphocytic leukaemia (CLL) the reaction ofthe lymphocytes for acid phosphatase is negative orweak, while in hairy cell leukaemia (HCL), a disorderwhich can be confused with CLL, the cells' reaction toacid phosphatase is moderately to strongly positiveand they are tartrate resistant. The enzyme ANAEgives similar results to acid phosphatase and has beencorrelated well with T cell markers by some,28 butthis reaction is particularly useful in characterisingthe monocytic component in a leukaemic marrow.The adaptation of plastic embedding to

undecalcified marrow cores provides excellent preser-vation of the marrow architecture and cellular detail.Even in such improved histological section prepara-tion, however, it is sometimes difficult to identify pre-cisely the nature of cellular diseases in the marrow,and enzyme or immunohistological studies becomeessential to characterise the type of cells or sanctuaryof cells participating in the malignant process. Thiscan now be done by embedding undecalcified marrowcores in GMA. Although water miscibility and lowtemperature processing make GMA a suitableembedding medium for enzyme histochemistry, thenature and type of fixative used is vitally important.We tested various fixatives and the general conclusionis that PGA and Bouin's solution best preserve theenzyme activity of marrow cells in semithin sectionsof undecalcified marrow cores processed in glycolmethacrylate.

Part of this work was carried out by one of us (AI)while working at the Medical Research CouncilLeukaemia Unit, Hammersmith Hospital, London.

References

1 Flandrin G, Bernard J. Cytological classification of acute leu-kaemias: a survey of 1400 cases. Blood cells 1975;1:7-15.

2 Kolakowska-Polubiec K, Litwin J. Usefulness of cytochemicalinvestigations in differential diagnosis and prognosis of acuteleukaemias. Mater Med Pol 1979;11:125-31.

3 Burkhardt R. Bone marrow and bone tissue. In: Colour atlas ofclinical histopathology. Berlin: Springer-Verlag 1971.

4 Block M. Bone marrow examination. Aspiration or core biopsy,smear or section, hematoxylin-eosin or Romanowsky stain-which combination? Arch Pathol Lab Med 1976;100:454-6.

5 Islam A, Frisch B. Plastic embedding in routine histology. I. Sec-tion preparation from undecalcified marrow cores. Histo-pathology 1985;9:1263-74.

6 Ruddell CL. Hydroxy-ethyl-methacrylate combined with poly-ethylene glycol 400 and water an embedding medium for rou-tine 1-2pm sectioning. Stain Tecnol 1967;42:119-23.

7 Ashford AE, Allaway WG, McCully ME. Low temperatureembedding in glycolmethacrylate for enzyme histochemistry inplant and animal tissues. J Histochem Cytochem1972;20:986-90.

8 Hoshino M, Kobayashi H. The use of glycolmethacrylate as anembedding medium for the histochemical demonstration ofacid phosphatase activity. J Histochem Cytochem1971;19:575-7.

9 Beckstead JH, Bainton D. Enzyme histochemistry on bone mar-row biopsies: reaction useful in the differential diagnosis ofleukemia and lymphoma applied to 2-micron plastic section.Blood 1980;55:386-96.

10 Beckstead JH, Halverson PS, Rie CA, Bainton DF. Enzyme his-tochemistry and immunohistochemistry on biopsy specimensof pathologic human bone marrow. Blood 1981;57:1088-98.

11 Jamshidi K, Swaim WR. Bone marrow biopsy and unalteredarchitecture: a new biopsy device. J Lab Clin Med1971 ;77:335-42.

12 Islam A. A new bone marrow biopsy needle with core securingdevice. J Clin Pathol 1982;35:359-64.

13 Hanker JS, Laszlo J, Moore JG. The light microscopic demon-stration of hydroperoxidase-positive Phi bodies and rods inleukocytes in acute myeloid leukaemia. Histochemistry1978;58:241-52.

copyright. on A

pril 21, 2021 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.40.10.1194 on 1 O

ctober 1987. Dow

nloaded from

1200 Islam, Henderson14 Goldberg AF, Baraka T. Acidphosphatase activity in human

blood cells. Nature (Lond) 1962;195:290-7.15 Yam LT, Li CY, Crosby WH. Cytochemical identification of

monocytes and granulocytes. Am J Clin Pathol1971;S5:283-90.

16 Leduce EH, Bernard W. Recent modifications of the gly-colmethacrylate embedding procedure. J Ultrastruct Res1967;19:196-9.

17 Spaur RC, Moriarty GC. Improvements of glycol-methacrylate.I. Its use as an embedding medium for electron microscopicstudies. J Histochem Cytochem 1977;25:163-74.

18 Sims B. A simple method of preparing 12pm sections of largetissue blocks .using flycol-methacrylate. J Micros1974;101:223-7.

19 Lee RL. 2-hydroxyethylmethacrylate embedded tissue-amethod complimentary to routine paraffin embedding. MedLab Sci 1977;34:231-9.

20 Westen H, Muck KF, Post L. Enzyme histochemistry on bonemarrow sections after embedding in methacrylate at low tem-perature. Histochemistry 1981;70:95-9.

21 Germain J. Epoxy resin embedding and light microscopy; itsadvantages and disadvantages. Science Tools, the LKB Instru-ment Journal 1974;21:30-3.

22 Islam A. Bone marrow structure in human leukaemias: a histo-

logical study by plastic embedding technique. London: Univer-sity of London, (PhD thesis.) 1982.

23 Romeis B. Mikroskopische Technik. Vienna: Oldenbourg R, 1974.24 Islam A. Haemopoietic stem cell: a new concept. Leuk Res

1985;9:1415-32.25 Islam A. Pattern of bone marrow regeneration following chemo-

therapy for acute myeloid leukaemia. Journal of Medicine (inpress).

26 Catovsky D, Galetto J, Okos A, Miliani E, Galton DAG. Cyto-chemical profile of B and T leukaemic lymphocytes with specialreference to acute lymphoblastic leukaemia. J Clin Pathol1974;27:767-7 1.

27 Brouet JC, Valensi F, Daniel MT, Flandrin G, Preud'homme JL,Seligman M. Immunological classification of acute lympho-blastic leukemia: evaluation of its clinical significance in a hun-dred patients. Br J Haematol 1976;33:319-28.

28 Higgy KE, Burns GF, Hayhoe FGJ. Discrimination of B, T andnull lymphocytes by esterase cytochemistry. J Haematol1977;18:437-48.

Requests for reprints to: Dr Anwarul Islam, Roswell ParkMemorial Institute, 666 Elm Street, Buffalo, New York14263 USA.

copyright. on A

pril 21, 2021 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.40.10.1194 on 1 O

ctober 1987. Dow

nloaded from