METHODS

Tissue samples

Needle muscle biopsies were obtained from the vastus lateralis of 20 normal healthy adults (16 men: age range, 16–58 years; mean age, 32 and four women: age range, 26–53 years; mean age, 36), frozen, and then processed as described previously.¹⁵ Conditions conformed to an ethical protocol granted by the Western Infirmary, Glasgow, Scotland.

Other material consisted of biopsy samples from patients with AIDS, which were formalin fixed and paraffin wax embedded, and postmortem samples from patients with a variety of neuromuscular diseases.

ATPase protocol

The standard ATPase method was used at pH values of 9.4, 4.6, and 4.3 on the frozen tissue samples, together with a negative control.¹⁵ For each case, two 10 μm serial sections were used, so that a total of eight was required. The ATPase preparations were done as part of the usual routine series of muscle stains¹⁵ (haematoxylin and eosin, modified Gomori trichrome, oil red O, periodic acid Schiff with and without diastase; and histochemistry for the demonstration of succinic and nicotinamide dehydrogenases, cytochrome oxidase, acid phosphatase, and the identification of deficiencies in myophosphorylase, adenylate deaminase, and phosphofructokinase).

IHC protocol

Two 10 μm sections from each case were placed on one slide. The control consisted of two sections of normal unrelated skeletal muscle, incorporated into each run of up to 10 cases. Commercial antibodies to fast and slow isoforms of myosin were used, each with a different visualisation system, so that specific identification of each fibre type on the same section was feasible. In brief, after fixation in acetone the antibody to slow myosin was applied, followed by a peroxidase conjugated rabbit antimouse antibody; the result was visualised as black type 1 fibres using the commercial Vector SG substrate kit (Vector Laboratories, Peterborough, UK). The alkaline phosphatase conjugated antibody to fast myosin was then applied; red type 2 fibres were visualised using the commercial Vector red substrate kit. The full details of the simple standard protocol¹⁶ (fig 1), used widely in routine pathology laboratories and which we modified slightly, are as follows.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Diagram of the staining protocol.

Sections were incubated with 20% normal rabbit serum (NRS) in Tris buffered saline pH 7.6 (TBS) for 10 minutes, after which the excess serum was drained off and the sections incubated in monoclonal antibody to slow myosin (Sigma-Aldrich, Poole, Dorset, UK) diluted 1/2000 in 4% NRS for 30 minutes, followed by three washes in TBS. The sections were then incubated in peroxidase conjugated rabbit antimouse antibody (Dako Ltd, Ely, Cambridgeshire, UK), diluted 1/50 in 4% NRS, for 60 minutes, and then washed three times in TBS. Vector SG peroxidase substrate solution (Vector Laboratories) was then applied, controlling the reaction by microscopic examination over five to 15 minutes. The sections were washed in running tap water, incubated in TBS for five minutes, and then in 20% NRS for 10 minutes. Excess serum was drained off before incubation in alkaline phosphatase conjugated monoclonal antibody to fast myosin (Sigma-Aldrich), diluted 1/50 in 4% NRS, for 60 minutes. After washing as above, the sections were incubated in Vector red alkaline phosphatase substrate solution (Vector Laboratories) for 10–20 minutes, controlling the reaction by microscopic examination. The final wash was in running tap water, after which the sections were dehydrated through graded alcohols, cleared in xylene, and mounted in synthetic medium.

The antibodies used were: (1) for type 1 fibres, monoclonal antimyosin (skeletal, slow; clone NOQ7.5.4D; Sigma-Aldrich). (2) For type 2 fibres, monoclonal antimyosin (skeletal, fast; alkaline phosphatase conjugate; clone MY-32; Sigma-Aldrich).

On completion, type 1 fibres are black whereas the type 2 fibres are pink. Type 2a and 2b subtypes can be distinguished because 2b are completely pink whereas 2a are intermediate between black and pink, appearing as a granular, dark, pinkish grey (fig 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Comparison of frozen sections stained for myofibrillar ATPase at (A) pH 9.4, (B) pH 4.6, and (C) pH 4.3 and (D) by immunohistochemistry for slow (dark grey) and pink (fast) myosin. Type 1 fibres are dark grey whereas type 2a and 2b fibres are distinguishable as pale greyish pink and pink, respectively. Original magnification, ×100.

The above method also works very well on formalin fixed, paraffin wax embedded tissue, following the usual preliminary steps—that is, sections taken to water and endogenous peroxidase removal with 0.3% hydrogen peroxide for 10 minutes. Antigen retrieval is necessary and this was achieved by five minutes at full pressure in a microwaveable pressure cooker (1 mM EDTA), followed by a 20 minute cool down period, one wash in water, and then trypsin digestion (0.1% trypsin/calcium chloride) for one minute. The method continues as above.

MORPHOMETRIC ANALYSIS

The SCION package (SCION Corporation, Maryland, USA) running on Microsoft Windows 95 and downloaded from the website http://www.scioncorp.com was used. It is capable of the wide range of image processing and analysis functions required for fibre analysis.^17–20 Linear and geometric measurements including diameter, length, area, and perimeter of fibres can be made. Up to 500 measurements can be stored at one time in the results window. Spatial calibration is supported by this program so it can provide real world area and length measurements (such as μm). A semiautomatic analysis routine is followed because this allows the observer to reject tissue artefacts.²¹

Tissue sections were viewed using the ×10 objective and visualised on the monitor. Several still images for each case were imported into the image analysis package, SCION Image for Windows, and morphometric data on muscle fibres retrieved by means of a semiautomated routine using the SCION software.

The following parameters were recorded:

1. Area: area of fibre cross section.

2. Mean density: average grey value within a selection.

3. X-Y centre: centre of the best fitting ellipse.

4. Modal density: most frequently occurring grey value within a selection.

5. Perimeter/length: length around the outside of a polygon selection.

6. Ellipse major axis: length of the major axis of the best fitting ellipse.

7. Ellipse minor axis: length of the minor axis of the best fitting ellipse.

8. Integrated density: sum of the grey values in a selection, with background subtracted.

9. Min/Max: minimum and maximum grey values within a selection.

Eighty fibres were counted at random on each section examined. Using the ATPase method, the 80 fibres were counted on the pH 9.4 preparation as type 1 and 2 and then a further 80 on the pH 4.6 section were analysed as type 2a or 2b. Using the IHC method, the 80 fibres were counted and classified on one preparation as type 1, 2a, or 2b.

Statistical methods

The staining methods were compared by analysis of variance to see whether they had different influences on estimates of the morphometric variables, such as cross sectional area and minor axis of the fibre cross sections.

To test whether the two staining methods gave different estimates of the proportions of type 1 and type 2 fibres, the χ² test was used. A separate χ² was calculated for each individual and the verdict was based on summation of the χ² components and degrees of freedom.

RESULTS

The 20 biopsies from the normal healthy controls yielded suitable samples. The morphometric and densitometric measurements had distributions sufficiently close to normal to allow the use of parametric statistical analysis.

It can be seen from table 1 that there was no significant difference (p = 0.57) between the overall percentages of type 1 and 2 fibres identified by the two different methods. For the group as a whole, the results were: type 1, 48% using the ATPase protocol v 46% using IHC and for type 2, 52% v 54%, respectively. When the groups were analysed as males and females, there were sex differences—men tended to have more type 2 fibres than women (55% and 56% v 42% and 45%, respectively) and women more type 1 fibres than men (58% and 55% v 45% and 44%, respectively), but again the results were similar for the ATPase and IHC methods.

Direct comparison of type 2 subtype percentages in the two methods was not possible. Both 2a and 2b fibres could be counted easily on the IHC sections. The subtypes could not, of course, be distinguished using ATPase at pH 9.4, but when the ATPase preparations at pH 4.6 and 4.3 were used, an overestimate of type 1 fibres always resulted. However, it can be seen from table 2 that the IHC subtype results we obtained are comparable to accepted values for vasti muscles.^{1–2, 7, 17, 22, 23}

Comparative analysis of cross sectional area (μm²) and diameter was carried out (table 3). Cross sectional area is regarded as being sensitive to the angle at which a fibre is cut and therefore the minor axis length (diameter), which is relatively insensitive to this variation, is used also. No significant differences were found. The IHC method made the fibres appear slightly smaller than the ATPase method, the minor axis being on average 2% smaller and the area 4% smaller. These differences were much less than the differences between individuals and they are caused by mild shrinkage occurring during the IHC method.

The form factor was also measured because it gives an indication of the number of fibres cut obliquely rather than transversely and therefore validates the above measurements.¹⁸ It is defined as 4πA/p², where A is the cross sectional area and p is the perimeter. A circle has a value of 1, whereas a straight line has a value of 0. Most normal muscle cells cut in true cross section give a form factor of 0.80 (range, 0.75–0.85). This was the case in both sets of sections.

Density of fibre staining was compared in the two methods, as a measure of substrate (enzyme or protein) concentration. Again, the values were comparable, with male type 2 fibres staining more than male type 1 and female type 1 and 2 being approximately equal on both protocols. The IHC sections could be used to analyse type 2 subtypes: this revealed that type 2b was more abundant than 2a in both sexes. The other parameters examined (X-Y centre, modal density, integrated density) also showed good correlation, and the fact that there were two colours made analysis easier.

The results given here show that there is excellent agreement between the traditional ATPase and the novel immunohistochemical method we propose. On archival and postmortem tissue the unique advantage of the new technique became apparent, with a clear and unambiguous distinction between the fibre types. Figure 3 illustrates this in a variety of conditions. In fig 3A, a case of neurogenic atrophy, fibres are of normal size in the upper half of the section but grouped and grossly atrophied in the lower part. Both type 1 and 2 fibres are seen in the atrophied area, helping to distinguish “primary” denervation from one occurring after reinnervation. In contrast, fig 3B/C is also from a case of neurogenic atrophy but does show evidence of reinnervation, with two fascicles composed almost entirely of one type or the other. In each fascicle, focal fibre atrophy is seen, indicating continuing denervation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

(A) Neurogenic atrophy. A group of normal sized fibres is seen in the upper part of the section with a severely atrophied group below. The latter includes both type 1 and 2 fibres, suggesting denervation before the establishment of reinnervation (original magnification, ×400). (B,C) Neurogenic atrophy (transverse section and longitudinal section, respectively) with reinnervation. Adjacent fascicles are composed almost entirely of one type or the other (original magnification, ×100). (D) Inflammatory myopathy. Normal mosaic pattern of type 1 and 2 fibres with severe selective atrophy of type 2. Architectural disorganisation (arrow) can be seen in a type 1 fibre (original magnification, ×200). (E) Severe wasting of type 2 fibres (original magnification, ×200). (F) Gross variation in size of both type 1 and 2 fibres in a case of muscular dystrophy. Distinction between the two types is clear (original magnification, ×400).

Figure 3D is from a case of polymyositis. The mosaic pattern of fibre types is preserved but there is obviously selective atrophy of type 2 fibres. Severe damage to myosin is detectable in a type 1 fibre. Muscle wasting associated with chronic obstructive airways disease is shown to result mainly from atrophy of type 2 fibres in fig 3E. Figure 3F illustrates the features in a case of muscular dystrophy. Although fibre type distinction can be blurred on the ATPase stain, it is clear on the IHC method, where gross variation in both fibre types, with occasional split ones, is detectable.

DISCUSSION

It is well established that human skeletal muscle consists of two main fibre types, divided according to metabolic and physiological features. Changes in the normal mosaic pattern of type 1 and 2 fibres—whether they show a proportionate increase or decrease, are grouped or not, hypertrophied or atrophied—are key features in the diagnosis of myopathies.^{1, 2} Fibre type changes in association with major cardiovascular risk factors (insulin resistance, hypertension, and obesity), chronic heart failure, and respiratory failure are also of current interest,^24–27 as is their relation to physical activity²⁸ and the selection and training of young athletes.^29–31

The myosin protein and the myosin ATPase enzyme are associated intimately both anatomically and functionally because myosin is the protein upon which ATPase acts to convert chemical energy to mechanical energy. Just as there are several isoforms of ATPase, allowing fibre type delineation, there are distinct isoforms of myosin heavy chains—at least seven have been identified in normal skeletal muscle, correlating with phenotypic differences in the contractile, metabolic, and size properties of fibres.¹⁴ Numerous investigations using histochemical, immunohistochemical, biochemical, and electrophoretic methods on normal skeletal muscle, including individual fibres, have confirmed that there is a strong correlation between ATPase activity and myosin heavy chain type.^{10–14, 32} Indeed, because many metabolic properties are only loosely coupled with the molecular properties of myofibrillar ATPases, and correlate better with myosin heavy chain isoforms, it has been stated categorically that: “ . . .myosin heavy chain isoforms appear to be the best choice for fibre delineation”.¹³

A histochemical reaction for ATPase, carried out with preincubation at different pH values, forms the basis on which fibres are classified as type 1 (slow, oxidative), 2a (fast, oxidative/glycolytic), and 2b (fast, glycolytic),^{1–8, 15} and normal values were established 30 years ago.^{6, 7, 15, 22, 23, 33–35} The method, however, has several disadvantages, including difficulty in achieving reproducibility because of the critical dependence on pH, temperature, and incubation time, in addition to the fact that the preparation fades and, of course, the problem that it will not work on fixed tissue because of enzyme degradation.

Our aim was to provide a simple and rapid IHC protocol, based on the identification of slow and fast myosin, as an alternative. We used commercially available materials. It would have been ideal to have each monoclonal antibody directly conjugated but unfortunately only one—alkaline phosphatase conjugated monoclonal antibody to fast myosin—could be obtained. If a peroxidase conjugated monoclonal antibody to slow myosin had also been available, we could have used a sequential direct method, instead of a combination of direct and indirect techniques. No doubt this modification will become available in the near future. However, the technique worked very well. The indirect method was performed first to avoid any detection of the conjugated antibody by the antimouse secondary layer. There was no advantage to using a biotinylated antibody in the first layer because no increase in sensitivity was needed. The Vector SG substrate gives a black end product, which incorporates the company's own enhancing effect and provides a highly effective contrast to the other (red) fibres much better than the usual brown colour produced when diaminobenzene is used.

“Changes in the normal mosaic pattern of type 1 and 2 fibres—whether they show a proportionate increase or decrease, are grouped or not, hypertrophied or atrophied—are key features in the diagnosis of myopathies”

The most important advantage of the IHC protocol is probably that it can be used in the diagnostic distinction of fibre types in paraffin wax embedded muscle tissue. Its other advantages include the ability to study both fibre types on the same preparation, economy of use, and the production of permanent and colourful preparations so that image analysis is easy. Economy of tissue use is important now that needle biopsies are becoming routine and micromethods for all parameters, including gene analysis, are available. An immunofluorescent technique was also considered but this would offer only a temporary preparation, which is unsuitable for morphometry, and examination is tedious. A few previous attempts at IHC protocols, using monoclonal antibodies to various neural^36–38 and muscle^{39, 40} antigens have been made. Two groups used antibodies to myosin but a single labelling technique was used, which could not identify type 2 subtypes.^{24, 40, 41} None is as satisfactory as the one described here.

The superiority of an IHC method that can be applied to paraffin wax embedded muscle has been stressed recently by workers using antibodies to dystrophin and the dystrophin associated glycoproteins in the diagnosis of muscular dystrophy.⁴² They pointed out that such a method obviates all the problems of freezing, handling, safety issues, and the storage of frozen material, in addition to allowing satisfactory analysis of archival tissue.

Our results indicated no significant differences in type 1 or 2 proportions and a clear distinction between type 2a and 2b with similar morphometry (cross sectional area, minor axis length, and form factor measurements). The minor axis was, on average, 2% smaller using IHC than with the ATPase method, and the cross sectional area was 4% smaller, probably because of a small amount of shrinkage occurring during the IHC method, but these differences were less than the differences between individuals. They make no significant difference to the interpretation of myopathy and are simply a minor factor to be considered in the learning curve of the method, as stated previously.⁴² The density of fibre staining was comparable, indicating a correlation between enzyme (ATPase) and protein (myosin) concentrations.

Myosin isoforms are regulated developmentally in the same way as myofibrillar ATPases. Using our IHC method, we could not distinguish between type 1 and the minor type 2C subtype, which forms approximately 2–3% of normal fibres and is considered to be a precursor of type 1. These fibres, however, are important only in the context of regeneration, when other features will be present. The IHC method will score in single fibre analysis when there may be hybrid expression of both myofibrillar ATPase and myosin heavy chain isoforms, but the former is very difficult to evaluate by staining intensity.¹²

As can be seen from the illustrations, the colour combination made analysis easy in the pathological diagnosis of both fixed and frozen tissue.

We conclude that this novel method offers a great improvement over the standard ATPase method.