Ultra-thin Sections of Avian Tubercle Bacilli in a 

 New Embedding Medium 



A.M.Glaueri' and E.M.Brihger 



Slrangcways Rcscarcli Lahuiatoiy (iiul Papworlli llospilul. Camhriclgc 



The reproduction of the tubercle bacillus presents 

 a particularly interesting subject for study. The 

 mycobacteria may be considered to occupy an 

 intermediate position between the ordinary bacteria, 

 which reproduce by binary tission, and the Nocardia 

 and fungus-like organisms that undergo a more 

 complicated life-cycle involving budding, branching 

 and mycelial development. 



We have previously studied the development of 

 the avian tubercle bacillus in the phase contrast 

 microscope (2) and were able to show that it can 

 reproduce in more than one way. In some strains 

 the rods of the original inoculum elongate and then 

 divide directly by binary fission; these organisms 

 behave like the ordinary bacteria and such strains 

 have been described as "bacillary" strains. In other 

 strains the original rods grow and branch, without 

 immediate division, to form complex mycelial struc- 

 tures. These mycelia finally break down by a process 

 of simultaneous multiple division to produce a mass 

 of rods similar in appearance to those of the original 

 inoculum. 



These two types of development of avian tubercle 

 bacilli have also been investigated in preparations of 

 intact organisms in the electron microscope (I). 



Only a limited amount of information can be 

 obtained from such a study owing to the thickness 

 of the bacilli, but it was obvious that a marked 

 difference in internal structure underlay the differen- 

 ces in the reproductive processes. At the same time 

 an unexpected development of round bodies or 

 "intracellular units'' was observed within the elon- 

 gated rods of the bacillary strains. These bodies have 

 since been examined more closely by the technique 

 of ultra-thin sectioning and the accompanying cyto- 

 plasmic changes have been observed. 



As other workers with bacteria have also observed, 

 we found that the organisms were frequently dis- 

 torted during the familiar preparative procedures for 

 ultra-thin sectioning. 



These observations led us to consider the possi- 

 bility of using an alternative embedding medium. 

 Maaloe and Birch-Andersen (5) have already had 

 considerable success with a resin of the epoxy type 

 and we have been experimenting with resins of the 

 same series.- The ones we have been investigating 

 are marketed under the trade name of Araldite. 



1 Sir Halley Stewart Research Fellow. 



- These resins were developed by Messrs. Ciba Ltd. ol 

 Basel and are made in England by Aero Research Ltd. of 

 Duxford, Cambridge. 



A Standard liquid epoxy resin is used and is made 

 more plastic by the addition of dibutyl phthalate. 

 On the addition of a suitable hardener the resin 

 sets uniformly, without shrinkage, to form a clear, 

 light-gold block. Originally we used an aliphatic 

 polyamine as the hardener, this being the usual type 

 of cold-setting hardener for these resins. Unfortu- 

 nately, as was discovered by Maaloe and Birch- 

 Andersen (5), a mixture of the resin with this hard- 

 ener is difficult to handle and is not ideal for electron 

 microscopy. The mixture is very viscous so that im- 

 pregnation is difficult and also it is not readily soluble 

 in absolute alcohol. On the advice of Dr. Glauert, of 

 Aero Research Ltd., we have experimented with a 

 different hardener with promising results. If the 

 casting resin is mixed with equal quantities of this 

 new hardener, which is a liquid anhydride, the 

 resultant mixture is sufficiently fluid to be handled 

 easily and is readily soluble in absolute alcohol. 

 The hardening process takes a considerable time at 

 normal incubation temperatures, but it can be 

 speeded up as much as required by the addition of an 

 amine accelerator. 



After various trials we have found the following 

 mixture to be suitable as an embedding medium: — 



^^ Araldite'' for iiltra-thiii sections 



Casting resin M 10.0 ml 



Hardener 964 B 10.0 ml 



Dibiit\ I phthalate 1.0 ml 



Accelerator 964 C 0.4 ml 



Slight variations of this formula ma\ be found con- 

 venient for difTcrcnt types of specimen. 



For ease of handling the resin is mixed and the 

 specimens are soaked at 48 C. The fixation o\' the 

 specimens in buffered osmic acid and dehydration 

 in graded alcohols is the same as for methacrylatc 

 embedding. From absolute alcohol the specimens 

 are passed to a 50 50 mixture of alcohol and Aral- 

 dite at 48 C for 1-2 hours, then to two changes of 

 the Araldite mixture at 48 C for 2-3 hours and 

 finally into gelatin capsules with fresh Araldite. The 

 blocks harden in about 30 hours at 48 C. It has been 

 found advisable to incubate the specimens at 48°C, 

 because at other temperatures there is a tendency 

 for soft specimens to rise in the cai-»siiic during the 

 setting process. 



The resultant blocks have a similar hardness to 

 methacrylatc and thin sections have been cut with 

 ease. Parallel specimens were embedded in metha- 

 crylatc so that a direct comparison could be made. 



