Fiber typing human skeletal muscle with fluorescent immunohistochemistry

Abstract

Skeletal muscle myosin heavy chain (MyHC) fiber type composition is a critical determinant of overall muscle function and health. Various approaches interrogate fiber type at the single cell, but the two most commonly utilized are single-muscle fiber sodium dodecyl sulfate-polyacrylamide gel electrophoresis (smfSDS-PAGE) and fluorescent immunohistochemistry (IHC). Although smfSDS-PAGE is generally considered the “gold standard,” IHC is more commonly used because of its time-effectiveness and relative ease. Unfortunately, there is lingering inconsistency on how best to accurately and quickly determine fiber type via IHC and an overall misunderstanding regarding pure fiber type proportions, specifically the abundance of fibers exclusively expressing highly glycolytic MyHC IIX in humans. We therefore 1) present information and data showing the low abundance of pure MyHC IIX muscle fibers in healthy human skeletal muscle and 2) leverage this information to provide straightforward protocols that are informed by human biology and employ inexpensive, easily attainable antibodies for the accurate determination of fiber type.

BRIEF BACKGROUND ON HUMAN SKELETAL MUSCLE FIBER TYPE

There are three primary myosin heavy chain (MyHC) proteins expressed in human skeletal muscle that determine fiber type: MyHC I, MyHC IIA, and MyHC IIX. In sequential order, each fiber type possesses distinct contractile characteristics ranging from slow (less powerful) to fast (more powerful) and metabolic profiles ranging from oxidative to glycolytic. Fiber type composition can exist along a continuum, where fibers express a single MyHC isoform or coexpress multiple MyHC isoforms (MyHC I/IIA, IIA/IIX, and I/IIA/IIX), yielding six different MyHC fiber isotypes (6, 26). In healthy individuals, roughly equal proportions of MyHC I and IIA fibers are present in biopsy samples of vastus lateralis and lateral gastrocnemius muscle, the two most abundant and commonly studied human muscles. However, this distribution is not static and can be altered by muscle use patterns (9). For instance, endurance athletes generally possess more MyHC I fibers (9, 24, 32, 33, 37), which is likely a consequence of the training stimulus (9, 29, 48, 55). Conversely, MyHC IIA fibers may become more prevalent in athletes participating in strength/power activities (5, 7, 25, 38, 47, 59). The appearance of “hybrid” fibers that coexpress MyHC IIA and IIX and a general shift to a more fast-twitch contractile phenotype often accompany sedentary behavior (20, 27, 59). A transition from comparatively more oxidative MyHC I and IIA fibers toward glycolytic coexpressing IIA/IIX fibers could characterize elite strength and power performance (16, 47) but is generally linked to poor health outcomes (14, 20, 52); for example, the abundance of IIA/IIX hybrid fibers is inversely correlated with aerobic fitness (19). An extreme manifestation of muscle disuse in the form of spinal cord injury (SCI) leads to an accumulation of pure MyHC IIX fibers (3, 12, 35). Despite these examples of how fiber type can differ under various physiological and pathological conditions, there is currently no consensus on the relative abundance of pure MyHC IIX fibers in healthy human skeletal muscle. Our position is that this confusion largely arises from methodological constraints and not biological variability.

Fiber type proportions can provide insight into overall muscle health and function, so accurate classification is important for research purposes. The goal of this Cores of Reproducibility in Physiology (CORP) article is to 1) provide data and perspectives on the abundance of pure MyHC IIX muscle fibers in muscle samples from healthy humans and 2) outline detailed, validated, and simple fluorescent immunohistochemistry (IHC) approaches for accurately assessing skeletal muscle fiber type, with specific focus on assessing the prevalence of pure MyHC IIX fibers. This article serves to explain fiber typing discrepancies among laboratories and techniques and offers several approaches to fiber typing via IHC that cater to specific research questions based on human biology, thereby expanding on prior foundational work (11).

LACK OF CONSENSUS IN REPORTING ABUNDANCE OF MYHC IIX-EXPRESSING FIBERS IN HUMAN SKELETAL MUSCLE BIOPSIES

IHC involves the acquisition of thin muscle cross sections via cryosectioning, followed by incubation with isoform-specific MyHC antibodies and fiber type identification via fluorescent microscopy; this is the most commonly used modern approach for determining fiber type and can also afford the advantage of distinguishing anatomical distribution of fiber type as well as fiber size. In 2012, Bloemberg and Quadrilatero published a comprehensive IHC methods paper in which they tested a variety of antibody combinations and provided a recommendation for determining fiber type in human vastus lateralis muscle samples (11). Their recommendation was to use a cocktail of primary antibodies available from the Developmental Studies Hybridoma Bank (DSHB; University of Iowa) that detect MyHC I (BA-D5 or BA-F8), IIA (SC-71), and IIX (6H-1) fibers, followed by isotype-specific fluorescent-conjugated secondary antibodies. With this strategy, the authors reported that pure MyHC IIX fibers comprise 6.2 ± 1.6% of total fibers in young healthy human muscle samples (n = 7 subjects). In contrast, ongoing studies among our laboratories in which vastus lateralis muscle samples from a cohort of young healthy individuals (n = 22 subjects, 10 men and 12 women, age 25 ± 2 yr; Fig. 1A) were fiber typed with Bloemberg and Quadrilatero’s antibody guidelines showed very few pure MyHC IIX fibers (3 of 2,552 fibers analyzed, or 0.1%; Fig. 1B). Using a similar IHC approach, Fry et al. recently reported that pure MyHC IIX fibers were undetectable in healthy adults (19), whereas Roberts et al. reported up to 20% (44). This lack of consensus regarding MyHC IIX fiber type percentage in healthy human skeletal muscle via IHC is disconcerting, and may reflect variation in analytical approach rather than true biological variability.

Fig. 1.

Fluorescent immunohistochemistry (IHC; A and B) and single-muscle fiber sodium dodecyl sulfate-polyacrylamide gel electrophoresis (smfSDS-PAGE; C and D) for myosin heavy chain (MyHC) fiber typing in vastus lateralis biopsy samples from a cohort of healthy adult men and women (n = 22; 10 men/12 women). A: representative IHC image of MyHC I (green), MyHC IIA (red), and MyHC IIA/IIX (blue/red) fibers. Scale bar, 50 µm. B: quantification of muscle fiber type via IHC (2,552 total fibers), with gray box emphasizing low MyHC IIX abundance. C: representative image of a smfSDS-PAGE gel showing the continuum of MyHC fiber types. D: quantification of smfSDS-PAGE data from mechanically dissected single muscle fibers (876 total fibers) from the same 22 subjects analyzed in B, with gray box emphasizing low MyHC IIX abundance. Skeletal muscle biopsy samples were obtained under resting conditions, and smfSDS-PAGE was conducted as described by our laboratories in detail elsewhere (36). IHC was conducted in the Toth laboratory according to the methods described in the text, and images were captured at ×20 magnification. Std, standard.

The “gold standard” technique for determining fiber type proportion in skeletal muscle samples is sodium dodecyl sulfate-polyacrylamide gel electrophoresis on isolated single muscle fiber segments (smfSDS-PAGE). In brief, smfSDS-PAGE involves manual dissection of individual muscle fibers, denaturation of muscle proteins contained in those fibers, and then electrophoretic separation and direct visualization of MyHC proteins on a polyacrylamide gel via silver or Coomassie staining (10, 36, 52). Because a larger section of the muscle fiber is analyzed relative to IHC (e.g., 1–3 mm in length), this technique provides a measure that better reflects MyHC expression throughout the fiber. Using this approach (Fig. 1C), we assessed a sizable number of manually dissected muscle fibers from the same 22 subjects reported above and confirmed that <1% of these muscle fibers expressed pure MyHC IIX (6 of 876 fibers analyzed, or 0.6%; Fig. 1D). Our data are congruent with numerous other smfSDS-PAGE investigations (1, 9, 17, 20–22, 31, 33, 39, 41, 42, 55, 59, 61). To further validate the scarcity of pure MyHC IIX fibers in healthy adults, we pooled data on single-muscle fiber MyHC isoform expression from a number of published studies (13, 36, 53). This analysis included 38 healthy older individuals (18 men and 20 women, age 69 ± 6 yr) and 1,761 fibers (Table 1). The conclusion from these data was the same, <1% pure MyHC IIX fibers, consistent with the majority of studies that have evaluated fiber type via smfSDS-PAGE in older subjects (4, 23, 28–30, 49, 60) as well as a recent single-muscle fiber proteomics study (40). Taken together, these data argue strongly that pure MyHC IIX fibers are rare in healthy humans throughout a broad age range. This conclusion may not apply to certain pathological conditions, such as SCI patients, in whom a much higher proportion of pure MyHC IIX (>20%) has been reported (3, 12, 35). Additionally, elite athletes who compete in high-power sports/activities may have a genetic predisposition toward a greater abundance of pure MyHC IIX fibers, as detailed in a recent case report profiling an Olympic-caliber sprinter (56); however, other elite strength/power athletes lack pure MyHC IIX (47), indicating that the Olympic sprinter was a very unique case. The data from strength/power athletes also points to how IIA/IIX fibers may be highly functional and not exclusively a by-product of sedentary behavior (47), which is consistent with findings in other mammals that display high IIA/IIX proportions along with superior speed performance and fatigue resistance (15).

Table 1.

Fiber type via smfSDS-PAGE in healthy aged individuals

	MyHC
	I	I/IIA	IIA	IIA/IIX	IIX	I/IIA/IIX
No. of fibers (%)	1,094 (62.1)	66 (3.7)	434 (24.6)	143 (8.1)	7 (0.4)	17 (1.0)

Values are from n = 38 subjects. MyHC, myosin heavy chain; smfSDS-PAGE, single-muscle fiber sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

With the knowledge that SCI patients consistently present with pure MyHC IIX fibers, we validated the specificity of DSHB antibodies, using a vastus lateralis muscle sample from an individual who was 13 yr postinjury (American Spinal Cord Injury Association classification “A”). Using the DSHB antibody BF-35, which specifically recognizes MyHC I and IIA but not IIX (11, 50), we confirmed the appearance of frequent pure MyHC IIX fibers (labeled with 6H-1 in Fig. 2A and unlabeled in Fig. 2B). A serial section incubated with SC-71 shows that this antibody has some affinity for MyHC IIX protein in human samples (Fig. 2C), in agreement with previous reports using Western blots (8, 50). By sufficiently diluting SC-71, Bloemberg and Quadrilatero reduced the appearance of cross-reactivity and reported pure MyHC IIX fibers to be 5–10% of total fibers in healthy muscle samples (11), whereas essentially none is detected by smfSDS-PAGE (Fig. 1).

Fig. 2.

Immunohistochemistry (IHC) demonstrating specificity of myosin heavy chain (MyHC) antibodies and regional MyHC accumulation. A–C: representative images of sections from a muscle biopsy from a spinal cord-injured man (50 yr old, American Spinal Cord Injury Association classification “A”, 13 yr postinjury), taken at the same exposure. A and B: fiber typing using BF-35 (MyHC I and IIA, green) and 6H-1 (MyHC IIX, red) (A) compared with BF-35 only (B) on the same section. C: a serial section of the same region as in A and B showing fibers strongly (MyHC IIA) and weakly (MyHC IIX) positive for SC-71 (green), demonstrating the preferential affinity of the SC-71 antibody for MyHC IIA but weak affinity for MyHC IIX. D and E: representative images of a longitudinal muscle section from a young healthy man showing a muscle fiber that transitions from pure MyHC IIX [negative for BF-35, no fluorescence (black arrow in D), positive for 6H-1 (red arrow in E)] to MyHC IIA/IIX [positive for BF-35 (green arrow in D) and 6H-1 (green and red arrow in E)]. Scale bar, 300 µm. F and G: SC-71 applied to a muscle sample from a healthy young subject at 1:100 (F) and 1:500 (G) dilutions, illustrating the effect of antibody concentration on cross-reactivity (white arrows point to differential reactivity in the same fiber on serial sections). MyHC I is pink in D–G. Muscle tissue was obtained from the Bamman laboratory (A–C) and the University of Kentucky Center for Muscle Biology Tissue Bank (https://www.uky.edu/chs/muscle/bank), and IHC was conducted in the Peterson laboratory (D and E). Fiber borders (laminin) are in blue, and all images were captured at ×10 magnification.

What might account for the discrepancy in relative abundance of pure MyHC IIX fibers? One possibility is that this discrepancy reflects a fundamental limitation of all IHC approaches for fiber typing muscle samples, specifically, that only a very small length of the muscle is evaluated. In most instances IHC muscle fiber sections are 6–10 µm thick, and in mice, rats, rabbits, amphibians, and birds MyHC expression can vary along the length of a muscle fiber (18, 34, 43, 45, 46, 51, 62, 63). We speculated that this same pattern of expression could occur in human muscle fibers and may in part explain why IHC yields results discrepant from smfSDS-PAGE. By applying BF-35 and 6H-1 antibodies (along with the MyHC I-specific BA-D5 antibody) to longitudinal muscle sections from a young healthy subject, we observed that muscle fiber type may transform along the length of the fiber (Fig. 2, D and E), similar to reports in muscle from smaller mammals. Thus a single fiber segment from a human biopsy sample could present as a MyHC IIA/IIX hybrid via smfSDS-PAGE but could appear as pure MyHC IIA or IIX on cross section depending on where along the longitudinal axis of the fiber the cross section was obtained; indeed, a larger proportion of hybrid fibers manifested via smfSDS-PAGE compared with IHC in the data presented here (Fig. 1). Other factors associated with the analytical methods could also impact relative abundance of MyHC isotypes in muscle fibers via IHC. These factors include exposure time during imaging and contrast adjustments after imaging, whether analysis is performed by a trained technician or with automated software (58), the dilution/concentration at which the antibody is used, as well as the duration of antibody incubation. Figure 2, F and G, illustrate how different dilutions of SC-71 alter the degree of reactivity of this antibody with MyHC IIX; 1:100 results in noticeable fluorescence in a pure MyHC IIX fiber, whereas 1:500 on a serial section shows that same fiber is almost devoid of fluorescence. Antibody reactivity could be further manipulated by using undiluted supernatant antibodies that have very high concentrations or by incubating for a short period of time (e.g., 1 h) versus overnight (11, 19, 44). All of these factors combined can contribute to discrepant results between smfSDS-PAGE and IHC, further emphasizing the necessity of understanding the prevalence of pure MyHC IIX in human muscle and arriving at consensus methods for accurate fiber type identification.

RECOMMENDATIONS FOR ACCURATELY FIBER TYPING HUMAN SKELETAL MUSCLE VIA IHC

The method used for fiber typing human skeletal muscle samples must consider time, resources, capabilities, and the research question at hand. SmfSDS-PAGE is the gold standard technique, but it is time consuming and tedious, requires considerable resources, and is low throughput. For these reasons, IHC is the more desirable choice for fiber typing human muscle samples. In agreement with Bloemberg and Quadrilatero (11), one can utilize an antibody cocktail containing BA-D5 (or BA-F8), SC-71 (or A4.74, which yields essentially identical results at low concentrations), and 6H-1 from DSHB at the concentrations indicated in Table 2 to detect MyHC I, MyHC IIA, and MyHC IIX, respectively. The entire continuum of fiber types, including MyHC I/IIA and IIA/IIX hybrids, can be visualized with this approach, with the exception of pure MyHC IIX because of cross-reactivity of the SC-71 antibody. The result will nevertheless be consistent with smfSDS-PAGE, as we have demonstrated above, and we further suggest that MyHC IIA/IIX and MyHC IIX fibers be classified as MyHC IIAX/IIX (57); this prudent, conservative nomenclature accounts for the range of IIA and IIX proportions in hybrid fibers while acknowledging the low and method-dependent abundance of MyHC IIX. To reliably detect pure MyHC IIX on cross sections (e.g., if evaluating injured/diseased/unique populations or wanting to visualize segmental MyHC expression), we recommend cutting a serial section and using BF-35 (Fig. 3), which identifies MyHC I and IIA proteins. Alternatively, BF-35 can be substituted for SC-71 in the antibody cocktail, which only limits detection of MyHC I/IIA hybrid fibers. MyHC I/IIA is a comparatively lower-abundance fiber type (generally <10%; see Fig. 1 and Table 1 and Refs. 1, 2, 11, 54, 61), but if it is of interest it can be detected by using BA-D5 and SC-71 (see above). We do not recommend further diluting the MyHC IIA antibodies beyond 1:500, as this may incorrectly reflect a complete lack of reactivity with MyHC IIX (in our hands) and may also affect detection of MyHC IIA.

Table 2.

Antibodies for IHC fiber typing

Antibody	Dilution (Concentration)	Source	Notes
1° Antibodies to match smfSDS-PAGE
BA-D5 (MyHC I)	1:100 (3.05 µg/mL)	DSHB	αMs IgG2B
SC-71 (MyHC IIA)	1:500 (0.85 µg/mL)	DSHB	αMs IgG1
6H-1 (MyHC IIX)	1:1 in PBS (10.5 µg/mL)	DSHB	αMs IgM, supernatant
Laminin	1:100	Sigma-Aldrich	αRb IgG
1° Antibody to detect pure MyHC IIX
BF-35 (MyHC I and IIA)	1:100	DSHB	αMs IgG1
2° Antibodies
Fluor 647	1:200	Invitrogen A21242	GtαMs IgG2B
Fluor 488	1:200	Invitrogen A21121	GtαMs IgG1
Fluor 555	1:200	Invitrogen A21426	GtαMs IgGM
Fluor 350	1:100	Vector CI-1000	GtαRb IgG
Alternative 1° antibodies
BA-F8 (MyHC I)	1:100	DSHB	αMs IgG2B
A4.74 (MyHC IIA)*	1:500	DSHB	αMs IgG1

DSHB, Developmental Studies Hybridoma Bank; Gt, goat; IHC, immunohistochemistry; Ms, mouse; MyHC, myosin heavy chain; Rb, rabbit.

^*A4.74 may cross-react less with MyHC IIX relative to SC-71, as shown via Western blot (8, 26, 43) and IHC (43), but still has affinity for MyHC IIX, especially at higher concentrations.

Fig. 3.

Visual representations of recommended immunohistochemistry (IHC) fiber-typing protocols. Muscle fiber borders (laminin, blue) are shown on all images for illustrative purposes. Left: pink is myosin heavy chain (MyHC) I, green is MyHC IIA, and red is MyHC IIX. Right: green is MyHC I and/or IIA, and no fluorescence is MyHC IIX.

IHC METHODS FOR FIBER TYPING

The antibodies for fiber typing can be found in Table 2. To maintain consistency across laboratories, we suggest using concentrated antibodies (if available) at the recommended dilutions/concentrations; supernatant antibodies can be diluted to similar concentrations if necessary but are unfiltered and could introduce variability. The protocol is optimized for a four-channel fluorescent microscope but can be modified as needed if fewer channels are available. We also recommend using automated detection software, such as MyoVision (58), for quantification of fiber type distribution since this removes subjectivity and promotes consistency. As with any IHC protocol, the results are largely dependent on how well the tissue was cryopreserved and the quality of histological sectioning; freeze-damaged tissue and/or poor sectioning will yield unfavorable results.

Step 1

Allow muscle samples to equilibrate to the temperature of the cryostat (−20 to −23°C, 20–30 min). Mount muscle sample and cut 6- to 10-µm sections with a fresh blade.

Step 2

Collect each section on charged glass slides. Dry at room temperature for a minimum of 1 h (sections can be dried for up to 16 h and still be used for fiber typing). Use a PAP pen to circle the sections, and let the PAP dry for at least 15 min.

Note.

On each slide, cut one section specifically for use as a no-primary/secondary-only control, so that background fluorescence can be assessed and corrected at the time of imaging.

Note.

If antibody application commences before samples are dry, the sections tend to wrinkle and/or develop air bubbles, which affect image quality.

Note.

If sections were cut previously, dried, and stored at −20 to −80°C, slides should be dried again for at least an hour after thawing.

Step 3

Make a primary antibody cocktail containing BA-D5, SC-71, and laminin diluted into 6H-1 supernatant-PBS (1:1; can substitute BF-35 for SC-71 if pure MyHC IIX is of interest, such as when fiber typing SCI patients or sprint athletes). Apply to sections and incubate in a humidified slide box at room temperature for 1 h. Incubate no-primary control sections in PBS.

Step 4

Wash sections in a Coplin jar for 5 min three separate times at room temperature, using fresh PBS for each wash.

Step 5

Make a secondary antibody cocktail (all diluted in PBS), apply to sections, and incubate in a humidified slide box at room temperature for 1 h. Apply secondary antibodies to no-primary control sections at this time as well.

Step 6

Wash sections in a Coplin jar for 5 min three separate times at room temperature, using fresh PBS for each wash.

Step 7

Mount slides with VectaShield (long term) or glycerol-PBS (1:1; immediate imaging) under a glass coverslip (#1.5 thickness).

Step 8

When imaging, the no-primary control sections can be used to help inform exposure times for the sections that received primary antibody. The 6H-1 antibody (MyHC IIX) sometimes underperforms, especially in low-quality tissue, so using the amount of staining present in BA-D5+ (MyHC I) fibers can inform where to set the exposure time.

Technical Notes Regarding Reproducibility

Every laboratory conducting IHC likely has different equipment and may deploy certain stylistic approaches that are difficult to document. These discrepancies could produce different results and/or interpretation of results between laboratories. Thus, to provide a basis for comparison to the results presented above, we offer a few technical notes for consideration.

Regarding tissue handling, all three laboratories involved in this work mount their muscle samples in tragacanth gum or Optimal Cutting Temperature compound and freeze the tissue mounts with fresh liquid nitrogen-cooled isopentane; reusing isopentane dilutes it and results in poor tissue quality. The quality of cryosections is usually readily apparent upon visual inspection with a light microscope at the time of sectioning; poor tissue quality will manifest as indistinguishable muscle fiber borders and/or holes within muscle fibers. The best muscle sections are usually obtained when using a fresh blade. Exposure times used to obtain fluorescent images are highly dependent on the microscope itself and the usage history of the light source, so it is challenging to offer guidance in this area. With that said, the Peterson laboratory uses a Zeiss Axio Scope upright microscope with a DAPI/Hoechst/aminomethylcoumarin acetate filter (exciter 350/50, emitter 460/50), a FITC/green fluorescent protein filter (exciter 480/30, emitter 535/40), a tetramethylrhodamine isothiocyanate/Cy3 filter (exciter 545/30, emitter 610/75), and a Cy5 filter (exciter 620/60, emitter 700/75), and exposure times are typically in the 200–800 ms range; filters with broader or narrower excitation and emission spectra could yield slightly different results. ×10 or ×20 images, depending on the size of the section, are sufficient to obtain high-quality fiber type data. Images are preprocessed with ZEN software where background is manually corrected based on fluorescence in no-primary control samples.

CONCLUSIONS

Using the simple and rapid IHC protocol above, accurate fiber typing in human samples can be obtained that will be reproducible and comparable across laboratories. Our recommended methods have the potential to overestimate the abundance of pure MyHC IIX fibers compared with smfSDS-PAGE, perhaps because of regional differences in MyHC abundance along the fiber. However, as we demonstrate, this overestimation is likely quite small in healthy adults throughout a broad age range. Accordingly, our data support the conclusion that pure MyHC IIX fibers are relatively rare in healthy human skeletal muscle, and we recommend usage of the nomenclature MyHC IIAX/IIX to describe IIX-containing fibers. Our hope is that the methods presented in this CORP article will be widely adopted to advance further understanding of skeletal muscle phenotype in health and disease and facilitate comparison of data across research laboratories.

GRANTS

This work was supported by grants from the National Institutes of Health to K. A. Murach (AR-071753), C. A. Peterson (AR-060701 and AG-046920), M. S. Miller (AG-031303), T. W. Tourville (AR-066729), and M. J. Toth (AG-033547, HL-077418, AR-069199).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.A.P. and M.J.T. conceived and designed research; K.A.M., C.M.D., K.K., T.B.V., T.W.T., and M.S.M. performed experiments; K.A.M., C.M.D., K.K., T.B.V., T.W.T., M.S.M., M.M.B., C.A.P., and M.J.T. analyzed data; K.A.M., C.M.D., K.K., T.B.V., T.W.T., M.S.M., M.M.B., C.A.P., and M.J.T. interpreted results of experiments; K.A.M. prepared figures; K.A.M. and C.A.P. drafted manuscript; K.A.M., C.M.D., K.K., T.B.V., T.W.T., M.S.M., M.M.B., C.A.P., and M.J.T. edited and revised manuscript; K.A.M., C.M.D., K.K., T.B.V., T.W.T., M.S.M., M.M.B., C.A.P., and M.J.T. approved final version of manuscript.