Skip to main content
NCBI home page
Journal of Anatomy logoLink to Journal of Anatomy
. 2016 Nov 11;230(3):398–406. doi: 10.1111/joa.12564

Structural and functional characteristics of the thoracolumbar multifidus muscle in horses

J A García Liñeiro 1, G H Graziotti 2,✉, J M Rodríguez Menéndez 2, C M Ríos 2, N O Affricano 2, C L Victorica 2
PMCID: PMC5314397  PMID: 27861847

Abstract

The multifidus muscle fascicles of horses attach to vertebral spinous processes after crossing between one to six metameres. The fascicles within one or two metameres are difficult to distinguish in horses. A vertebral motion segment is anatomically formed by two adjacent vertebrae and the interposed soft tissue structures, and excessive mobility of a vertebral motion segment frequently causes osteoarthropathies in sport horses. The importance of the equine multifidus muscle as a vertebral motion segment stabilizer has been demonstrated; however, there is scant documentation of the structure and function of this muscle. By studying six sport horses postmortem, the normalized muscle fibre lengths of the the multifidus muscle attached to the thoracic (T)4, T9, T12, T17 and lumbar (L)3 vertebral motion segments were determined and the relative areas occupied by fibre types I, IIA and IIX were measured in the same muscles after immunohistochemical typying. The values for the normalized muscle fibre lengths and the relative areas were analysed as completely randomized blocks using an anova (P ≤ 0.05). The vertebral motion segments of the T4 vertebra include multifidus bundles extending between two and eight metameres; the vertebral motion segments of the T9, T12, T17 and L3 vertebrae contain fascicles extending between two and four metameres The muscle fibres with high normalized lengths that insert into the T4 (three and eight metameres) vertebral motion segment tend to have smaller physiological cross‐sectional areas, indicating their diminished capacity to generate isometric force. In contrast, the significantly decreased normalized muscle fibre lengths and the increased physiological cross‐sectional areas of the fascicles of three metameres with insertions on T9, T17, T12, L3 and the fascicles of four metameres with insertions on L3 increase their capacities to generate isometric muscle force and neutralize excessive movements of the vertebral segments with great mobility. There were no significant differences in the values of relative areas occupied by fibre types I, IIA and IIX. In considering the relative areas occupied by the fibre types in the multifidus muscle fascicles attached to each vertebral motion segment examined, the relative area occupied by the type I fibres was found to be significantly higher in the T4 vertebral motion segment than in the other segments. It can be concluded that the equine multifidus muscle in horses is an immunohistochemically homogeneous muscle with various architectural designs that have functional significance according to the vertebral motion segments considered. The results obtained in this study can serve as a basis for future research aimed at understanding the posture and dynamics of the equine spine.

Keywords: horse, multifidus muscle, muscle fibre type, spine

Introduction

The epaxial musculature is determined by muscle innervation provided by dorsal spinal nerve branches. These muscles are involved in rotational movements around the transverse (extension), vertical (laterality) and axial axes of the vertebral column. In addition, these muscles counteract passive movements induced by gravitational and inertial forces, and active movements generated by antagonistic muscles or transmitted to the trunk by extrinsic limb muscles, as Schilling & Carrier (2009) showed in canines. The hypaxial muscles, which are innervated by the ventral spinal nerve branches, are involved in flexion and the lateral movements of the vertebral column (Nickel et al. 1985; Barone, 1989; van Weeren, 2009; Hyytiäinen et al. 2014).

The transversospinalis muscle has a deep part consisting of the multifidus muscle, the caudal insertions of which occur on the lumbar mammillary processes, thoracic transverse processes and cervical articular processes until reaching the lamina (shortest fascicles) or the spinous processes (longest fascicles) and which pass through one or more metameres (M) (Macintosh et al. 1986; Barone, 1989). A metamere is a segment comprising two halves of successive vertebrae. A metamere includes a muscle segment, a divided bone segment (two successive vertebrae) and a segmentary nerve (Barone, 1989). The fascicles of the multifidus muscle are innervated by branches of the dorsal spinal nerve belonging to the metamere of their cranial insertion point (Macintosh et al. 1986).

In investigating the role of multifidus muscle in humans, Rosatelli et al. (2008) reported the presence of superficial fascicles as well as intermediate fascicles in the lumbar region. The former generate lumbar extension, whereas the latter control intersegmental movements. Furthermore, Moore et al. (2013) and Rosatelli et al. (2008) reported that the deep fascicles provide proprioceptive feedback. Moreover, the underlying controlling mechanism for this feedback was described by Panjabi (1992) who argued that the deformation of ligaments and tendons that occurs during movements generates proprioceptive information. This information produces the feedback that regulates the tensional activity of the muscles that stabilize the spine. In addition, Macintosh & Bogduk (1986) demonstrated that the forces produced by the human lumbar multifidus muscle could be resolved into a major vertical vector, favouring rotation about the transverse axis (extension).

Ritter et al. (2001), who considered the canine a model cursorial species, recorded ipsilateral and contralateral activities in the thoracolumbar multifidus muscle during the second half of the hind limb support phase. These authors also reported that this muscle diminishes flexion of the trunk. Additionally, Schilling & Carrier (2009) reported that the canine multifidus muscle performs heterogeneous functions because it produces rotations and stabilizes the spine against the gravitational and inertial forces generated by the trunk and the pelvic limbs.

Kaigle et al. (1995) reported that the multifidus muscle acts as a spinal stabilizer in pigs. However, the role of this muscle in other husbandry species has not investigated because this information would not increase knowledge regarding meat quality characteristics.

Although the structure of the vertebral column and the types of locomotion of humans and various cursorial species have been characterized, it is unacceptable to apply the findings for a model species to comparative studies (Bogduk et al. 1992; cited by Hansen et al. 2006).

In the thoracolumbar region of horses, multifidus bundles run craniomedially inserting on the lateral surface of the spinous processes of the cranial vertebra after crossing between 2M, which is called the long rotator muscle (R) to 3M (Barone, 1989), only 1M named the short rotator muscle to 5M (Stubbs et al. 2006; Hyytiäinen et al. 2014), 1M to 4M (Mc Gowan et al. 2007) or 2M to 6M (Nickel et al. 1985). The 2M fascicles are difficult to isolate in horses (Barone, 1989).

The vertebral motion segment (VMS), which is anatomically formed by two adjacent vertebrae and the interposed soft tissue structures (Haussler, 1999), is an important concept in this study. Although the range of motion of each segment is small, the sum of all of the segments results in considerable movements (Barone, 1989). Excessive mobility is well known to frequently cause osteoarthropathies in sport horses. Moreover, a relationship between equine spinal bone pathologies and asymmetrical cross‐sectional areas of the multifidus muscle has been demonstrated (van Weeren et al. 2010; Stubbs et al. 2011).

Townsend et al. (1983) demonstrated that the extent of movements resulting in flexion and extension of the equine spine are greater at the lumbosacral joint followed by the thoracic T1‐T2 VMSs and found that the majority of bending occurs at T11‐T12, whereas axial rotation occurs at T9‐T14. Denoix (1999) also reported that a greater degree of flexion occurs at the equine lumbosacral joint, followed by the T17‐L1 VMSs. Further studies conducted by Zaneb et al. (2013), who recorded spinal movements during the walk, trot and canter gaits, obtained results similar to those described by Townsend et al. (1983), except for the occurrence of axial rotation in the sacral region.

Stubbs et al. (2006, 2011), Mc Gowan et al. (2007) and Clayton et al. (2012) demonstrated the importance of the equine multifidus muscle as a VMS stabilizer. At least in humans, R appears to be activated prior to the superficial multifidus fascicles to counteract the rotation of the spine produced by the concentric contraction of the hypoaxial muscles (abdominal and sublumbar muscles; Moseley et al. 2002).

Recently, Hyytiäinen et al. (2014) reported that the biochemical structure of the equine multifidus muscle, i.e. the percentage of type I, IIA and IIX muscle fibres, corresponds to its function as a stabilizer of the spine. The architectural design of a muscle is very important in determining its functional parameters (Burkholder et al. 1994; Kearns et al. 2002; Felder et al. 2005; Payne et al. 2005; Eng et al. 2008). However, to our knowledge, no structural studies of the equine multifidus muscle have been conducted.

Considering the importance of the multifidus muscle to the posture and dynamics of the equine spine, its relationship with osteoarticular pathologies in sport horses (van Weeren et al. 2010; Stubbs et al. 2011) and the scarce documentation of the structure and function of this muscle (Schilling & Carrier, 2009), we performed a morphological study of the equine multifidus muscle in VMSs with different functionalities. We hypothesized that the characteristics of the multifidus muscle would vary according to the VMS considered.

Materials and methods

Gross anatomy

This research was performed on six carcasses of Silla Argentino breed horses, which were delivered to the Anatomy Department of the Veterinary Science School (University of Buenos Aires) no more than 24 h postmortem and showed no signs of locomotor system pathology. Two transverse cuts were made to obtain a segment of the thoracolumbar spine between the 6th cervical vertebra and the level of ilium bones. After removing the septum located between the erector spinae muscle and the deep transversospinalis muscle and disarticulating the sacroiliac joint to remove the ilium, the sample was immersed in a 10% formalin solution to preserve the physiological length of the muscles (Lieber & Fridén, 2000). Afterward, the multifidus muscle fascicles on the right side were detached from the cranial insertion on the spinous processes (Stubbs et al. 2006) to the caudal insertion (adjacent to the transverse process) via manual dissection. The muscles studied were those belonging to the VMSs corresponding to the T4, T9, T12, T17 and L3 cranial vertebrae. The architectural muscle design study was conducted according to an adaptation of the general rules of Hermanson (1997) (equine), Delp et al. (2001) (human), Graziotti et al. (2004) (llama) and Sharir et al. (2006) (canine) in accordance with the multifidus muscle measurements. In brief, the muscles were cleaned of fat and connective tissue, submerged in 10% formalin (for 7 days), rinsed with tap water (for 6 h) and then immersed in a 15% nitric acid solution (for approximately 3 days). When the fibre bundles easily could be teased apart, the muscle was rinsed overnight in a tap‐water drip, then immersed in a 0.4‐mol phosphate‐buffered solution of pH 7.2 (for 2 h) and finally submersed in 6% formalin at 4 °C until analysis. First, the distances between the aponeuroses that house the muscle fibres were measured at five different points in each muscle using a Vernier calliper (accuracy 0.01 mm). An average of 10 fibre bundles containing approximately 20 fibres were dissected from different sites of the muscle under a dissecting magnifying glass and the fibre lengths (FL) were measured using a calliper. To normalize the FL to the sarcomere length, the dissected fibres (10) of each muscle from one horse chosen at random were mounted using glycerol. A digital image of the sarcomere was captured (1000×, TIFF format), and the mean sarcomere length was obtained using scion image version β software. The mean normalized fibre lengths (NFL) were obtained using the following equation: FL × optimal horse sarcomere length/sarcomere length, in accordance with the method of Felder et al. (2005) and Eng et al. (2008), in which the optimal equine sarcomere length was 2.80 μm, in accordance with Marx et al. (2006). The pennation angle of a NFL was measured as follows: parallel lines (scale 10 : 1) at the average distances between the aponeuroses that house the muscle fibres were transcribed on a sheet of graph paper. Subsequently, a line (scale 10 : 1) representing the average NFL was traced between the lines representing the aponeuroses. Then, the pennation angle was measured using a goniometer. Afterward, the perimysium of every muscle was carefully separated under a magnifying glass, and the weight of the muscle fibres, the muscle mass (in grams), was determined using a digital electronic balance. The physiological cross‐sectional area (PCSA), measured in cm2, was calculated according to the following equation (Eng et al. 2008): muscle mass × cos θ/ρ × NFL, where θ was the pennation angle of the muscle fibres and ρ was the muscle density (1.056 g cm−3).

Immunohistochemistry, histochemistry and image analysis

Belly core samples (1.5 × 0.5 × 0.5 cm) of the multifidus muscles attached in the VMSs mentioned above were obtained from the left side of the same anatomical parts of the each horse. The samples were rolled in talcum powder, frozen by immersion in liquid nitrogen and maintained at −80 °C until analysis. The following assays were performed to identify the various fibre types: a myofibrillar adenosine triphosphatase (mATPase) assay, a nicotinamide adenine dinucleotide tetrazolium reductase (NADH‐TR) assay and immunohistochemical assays. Several 9‐μm‐thick serial cross‐sections obtained using a cryostat set at −27 °C were mounted on glass slides and subjected to the mATPase assay after acid preincubation (pH 4.3, for 5 min) (Brooke & Kaiser, 1970; modified by Nwoye et al. 1982). To evaluate the oxidative capacity, other serial sections were stained for NADH‐TR according to the method of Dubowitz & Sewry (2007). Additional serial sections were incubated with a panel of monoclonal antibodies that were specific for various isoforms of myosin heavy chain (Table 1). The immunohistochemical procedure resulting in an avidin‐biotin peroxidase complex was used to localize the bound primary monoclonal antibodies, performed as described by Graziotti et al. (2004). In brief, the sections were pre‐incubated in a blocking solution of stock goat serum. The primary monoclonal antibody, diluted in a saline‐containing 0.1 mol phosphate buffer, was applied to the sections for 40 min in a humid chamber at 37 °C. The sections were then washed and incubated with a secondary antibody. Then, the sections were washed and incubated with the avidin‐biotin peroxidase complex (Vectastain Elite ABC Kit standard PK‐6100, Vector Laboratories, Inc., Burlingame, CA, USA). Diaminobenzidine tetrahydrochloride was used as a chromogen to localize the peroxidase‐containing complexes. The proportion of hybrid fibres observed was considered irrelevant (< 3.5%). Thus, the immunohistochemically delineated fibre types were characterized as pure fibre types I, IIA or IIX (Table 1) according to Rivero et al. (1996) and Eizema et al. (2003). For quantitative analysis, the cross‐sectional areas occupied by the various fibre types were determined in images captured in TIFF format using motic image plus 2.0 software (Motic China Group Co., Ltd). The cross‐sectional area of each individual fibre type in the slides treated to visualize mATPase was determined from an image (100× magnification) by tracing the cell borders (of at least 100 fibres). Open source software (scion image version β) was utilized to determine the cross‐sectional areas, after which the relative areas (area occupied by each fibre type) were calculated.

Table 1.

Identification of fibre types according to specific staining with monoclonal antibodies (MAbs) directed against skeletal myosin heavy chain (MHC I, IIa, and IIx) isoforms (Rivero et al. 1996), mATPase reaction after preincubation at pH 4.3 and NADH‐TR reaction

Fibre types MAb slow BA‐F8 1 : 10 MAb A.474 1 : 10 mATPase (pH 4.3) NADH‐TR
I + − ++++ ++++
IIA − + − ++
IIX − ± ++ +
Open in a new tab

Statistical analysis

The NFL, PCSA, muscle mass and relative area values were analysed using an anova in a completely randomized block design in which the sampled animals were considered blocks (n = 6) (P ≤ 0.05). The significance of the differences were analysed using Tukey (NFL and relative area values) and Bonferroni (PCSA and muscle mass values) tests (P ≤ 0.05). In all cases, the normality of the values was tested using the Shapiro–Wilk test with a signification level of 10%. To achieve an acceptable power level (Ghasemi & Zahediasl, 2012) we set the significance level of the Shapiro–Wilk test at 10% to minimize the probability of type II errors. Because normality was observed for all of the analysed variables, only parametric tests were performed.

Results

Gross anatomy

Dissection of the multifidus muscles showed that the architectural design of all of the bundles comprised three sheets of fibrous connective tissue that act as a tendon or aponeurosis arranged in parallel (Fig. 1A), joining the mammillary or transverse processes with the spinous process of the vertebra corresponding to the cranial insertion point (T4, T9, T12, T17, and L3 VMSs; Fig. 2, Table 2). The fibre bundles, the NFL of which are indicated in Table 2, are located between the two parallel planes of the aponeuroses (Fig. 1, Fm). The deepest and most ventral fascicles belong to the R muscle, which we confirmed to occur in the thoracic and lumbar regions (Fig. 2). The VMS that is cranially indicated by the T4 vertebra contains bundles extending between two and eight metameres (RT4 to 8MT4), whereas the VMS including the T9, T12, T17 and L3 vertebrae contains fascicles that extend across two to four metameres (R‐3M‐4M). 5M fascicles inserted on T9 were found in two of the horses examined. Details regarding the metameres, including the respective insertion sites and NFL of the fascicles, are shown in Table 2. The mean NFL values of the fascicles included in each VMS (T4, T9, T12, T17, and L3) are shown in Table 3. The mean PCSA values for each muscle are provided in Table 6, and the mean PCSA values for the muscles within each VMS are shown in Table 7. The muscle mass values are displayed in Table 8. In brief, the NFL values varied significantly in descending order from 7MT4, 8MT4, 3MT4, 6MT4, 4MT9, 4MT17, 4MT4, 4MT12, RT4, 3MT9, 3MT17, RT9, 3MT12, RT12, 3ML3, 4ML3, RT17 to RL3. The PCSA values varied significantly in descending order from 3ML3, 3MT12, 3MT17, 4ML3, 4MT17, 4MT12, 4MT9, 3MT9, 4MT4, RT12, RT17, 6MT4, 7MT4, RT9, RT4, RL3, 3MT4 to 8MT4. The values of muscle mass were ordered significantly from major to minor: 7MT4, 4MT17, 4MT9, 3ML3, 4MT12, 6MT4, 3MT12, 4MT4, 3MT17, 3MT9, 4ML3, 3MT4, RT12, RT4, 8MT4, RT9, RT17, RL3.

Figure 1.

Figure 1

Open in a new tab

Typical structural design of an equine multifidus muscle, showing the aponeurotic sheets (A) and short pinnate muscle fibres (Fm).

Figure 2.

Figure 2

Open in a new tab

Equine multifidus muscle after removal of the erector spinae muscle. The image includes 2M–8M multifidus bundles that insert on T4; X indicates the transverse process of the T12 vertebra. The uncovered areas on the T2‐T10 spinous processes demonstrate the insertion of the spinalis muscle. Markers located at the apex of the spinous processes indicate the corresponding vertebra.

Table 2.

Composition of equine multifidus muscles according to their vertebral insertions, metameric extensions and NFL (cm). The caudal insertion points of the 3ML3 and 4ML3 fascicles involve five or six lumbar vertebrae. No muscle fibre length data were obtained for 5MT4. Values with the same superscripted letter are not significantly different (P ≤ 0.05). SE, standard error, df: 86

Multifidus Insertions Metameres NFL SE
7MT4 T4‐T11 7 2.84a 0.14
8MT4 T4‐T12 8 2.63ab 0.18
3MT4 T4‐T7 3 1.95bc 0.13
6MT4 T4‐T10 6 1.60cd 0.13
4MT9 T9‐T13 4 1.58cde 0.13
4MT17 T17‐L3 4 1.52cdef 0.13
4MT4 T4‐T8 4 1.40cdefg 0.14
4MT12 T12‐T16 4 1.28cdefgh 0.13
RT4 T4‐T6 2 1.18defgh 0.13
3MT9 T9‐T12 3 1.07defgh 0.13
3MT17 T17‐L2 3 0.95defgh 0.13
RT9 T9‐T11 2 0.90efgh 0.13
3MT12 T12‐T15 3 0.83fgh 0.13
RT12 T12‐T14 2 0.77gh 0.13
3ML3 L3‐L6/S1 3 0.74gh 0.13
4ML3 L3‐S1/S2 4 0.69h 0.13
RT17 T17‐L1 2 0.63h 0.13
RL3 L3‐L5 2 0.62h 0.13
5MT4 T4‐T9 5 – –
Open in a new tab

Table 3.

NFL (cm) of multifidus muscles that insert into the VMSs including the T4, T9, T12, T17 and L3 vertebrae. Values with the same superscripted letter are not significantly different (P ≤ 0.05). SE, standard error, df: 99

VMS NFL SE
T4 1.85a 0.09
T9 1.18b 0.12
T12 1.03b 0.12
T17 0.96b 0.12
L3 0.68b 0.12
Open in a new tab

Table 6.

PCSA values (cm2) for each of the muscles studied. Values with the same superscripted letter are not significantly different (P ≤ 0.05). SE, standard error, df: 90

Muscles PCSA SE
8MT4 0.88a 1.82
3MT4 2.26a 1.27
RL3 2.77a 1.27
RT4 3.05a 1.27
RT9 3.94ab 1.27
7MT4 4.31ab 1.40
6MT4 4.50ab 1.27
RT17 4.61ab 1.27
RT12 4.84ab 1.27
4MT4 5.67abc 1.40
3MT9 5.70abc 1.27
4MT9 6.24abc 1.27
4MT12 6.96abc 1.27
4MT17 7.14abc 1.27
4ML3 7.27abc 1.27
3MT17 7.65abc 1.27
3MT12 9.88bc 1.27
3ML3 12.29c 1.27
Open in a new tab

Table 7.

PCSA values (cm2) of the multifidus muscles that insert into the T4, T9, T12, T17 and L3 VMSs. Values with the same superscripted letter are not significantly different (P ≤ 0.05). SE, standard error, df: 103

VMS PCSA SE
T4 3.59a 0.65
T9 5.29ab 0.85
T17 6.47ab 0.85
T12 7.23b 0.85
L3 7.44b 0.85
Open in a new tab

Table 8.

Muscle mass (Mm) of each muscle. The values provided are the mean values of the original weights in grams. Values with the same superscripted letter are not significantly different (P ≤ 0.05). SE, standard error, df: 90

Muscles Mm SE
RL3 7.67a 6.66
RT17 14.33ab 6.66
RT9 27.50abc 6.66
8MT4 28.04abcd 9.54
RT4 30.33abcd 6.66
RT12 31.67abcd 6.66
3MT4 36.50abcd 6.66
4ML3 47.0bcde 6.66
3MT9 56.0cde 6.66
3MT17 61.50cde 6.66
4MT4 64.66cde 7.32
3MT12 65.33de 6.66
6MT4 68.0de 6.66
4MT12 77.67ef 6.66
3ML3 79.33ef 6.66
4MT9 81.17ef 6.66
4MT17 82.83ef 6.66
7MT4 84.26ef 7.32
Open in a new tab

Immunohistochemistry

Regarding the characteristics of the muscle fibres observed, type I and IIA fibres were identified in all of the muscles examined, whereas the phenotypic expression of type IIX fibres was not observed in some muscles, as shown in Fig. 3 and Table 4. There was no significant difference in the relative areas occupied by the type I, IIA and IIX fibres in the studied muscles. Considering the relative areas occupied by the various fibre types in the multifidus muscles attached to each VMS evaluated, the relative area occupied by type I fibres was significantly higher in VMS T4 (Table 5). There was no significant phenotypic expression of IIX fibres in the muscles attached to VMS T4.

Figure 3.

Figure 3

Open in a new tab

Serial cross‐sections of the equine rotator muscle inserted at T17, which is stained using enzyme histochemistry for myofibrillar ATPase after preincubation at pH of 4.3 (A) and for nicotine amide dinucleotide tetrazolium reductase (NADH‐TR) (D) and stained using immunohistochemistry with MAbs for slow (F‐8) (B) and A4.74 (C) MHC. The fibres labelled 1, 2, and 3 contain MHC I, MHC IIA and MHC IIX, respectively.

Table 4.

Mean relative areas (RA) occupied by type I, IIA and IIX fibres in the studied muscles. No significant differences in the values were found (P ≤ 0.05). df: 45 (I), 45 (IIA) and 24 (IIX)

Muscles RA I RA IIA RA IIX
RT4 0.548 0.413 –
3MT4 0.532 0.467 –
4MT4 0.611 0.386 0.003
6MT4 0.647 0.353 –
7MT4 0.670 0.329 –
8MT4 0.60 0.39 –
RT9 0.539 0.403 0.058
3MT9 0.47 0.465 0.065
4MT9 0.542 0.293 0.164
RT12 0.397 0.469 0.134
3MT12 0.468 0.404 0.128
4MT12 0.443 0.44 0.117
RT17 0.442 0.415 0.143
3MT17 0.456 0.378 0.166
4MT17 0.403 0.315 0.282
RL3 0.438 0.462 0.1
3ML3 0.454 0.479 0.067
4ML3 0.481 0.353 0.166
Open in a new tab

Table 5.

Relative areas occupied by type I fibres in the multifidus muscles that insert into the VMSs including the T4, T9, T12, T17 and L3 vertebrae. Values with the same superscripted letter are not significantly different (P ≤ 0.05). SE, standard error, df: 54

VMS RA I SE
T4 0.58a 0.02
T9 0.48ab 0.03
T12 0.44b 0.03
T17 0.44b 0.03
L3 0.43b 0.03
Open in a new tab

Discussion

Published descriptions of equine multifidus muscle with 2M and 3M (Barone, 1989) and 2M to 6M (Nickel et al. 1985) fascicles do not include details about the vertebral segments in which they are located. More recently, Stubbs et al. (2006, 2011) and Hyytiäinen et al. (2014) reported the presence of multifidus muscle with 1M to 5M fascicles along the thoracolumbar spine. Our results agree partially with those of these authors because we found 6M, 7M and 8M multifidus muscle fascicles that insert in the T4 VMS. In our study, we found 5M fascicles in the T9 VMS in 33% of the animals examined. In the T12, T17 and L3 VMSs, only 2M to 4M bundles were observed. Furthermore, the studies cited above describe 1M muscles between the T4‐L6 segments, which originate and insert into the same metamere (short rotator muscle). We disagree with these authors because we found a muscle fascicle inserting into the same T4‐T5 metamere in only one animal on the right side. It is not known whether the breed, sex or age of an individual is associated with the presence of such a muscle. Burkholder et al. (1994) and Eng et al. (2008) have argued that muscular structures are conserved among species, even those that have very different patterns of locomotion.

Fibre length is an important feature of muscular design. According to Burkholder et al. (1994), Kearns et al. (2002), Felder et al. (2005), Payne et al. (2005) and Eng et al. (2008), fibre length is directly proportional to the maximal shortening velocity of a muscle and is inversely proportional to its ability to generate isometric force. Differences in maximal speed depend more on shortening the length of the muscle fibre than on its biochemical characteristics (Burkholder et al. 1994; Kearns et al. 2002). Although Delp et al. (2001) and Sharir et al. (2006) considered the changes in the length of the fascicles due to the treatment employed in the present study to be insignificant (< 3%), we decided to make our measurements more accurate by normalizing the fibre length to the sarcomere length, consistent with Lieber & Fridén (2000), Felder et al. (2005) and Sharir et al. (2006), considering that sarcomere length and fibre length are linearly related.

Robert et al. (2001) demonstrated that the longissimus thoracolumbar muscle stabilizes the spine by putting pressure on the vertebrae, making that region an efficient platform for transmitting propulsion impulses from the hind limb. However, the kinematics of the cranial end of the platform (withers) depend more on the activity of the forelimb, head and neck movements. Additionally, Licka et al. (2009) argued that the longissimus muscle acts as a stabilizer of the spine, reaching its highest activity preventively when the momentum of the ipsilateral hind limb (at the walk) occurs, with the major activity occurring at T12. In that context, the presence of the longest 5M, 6M, 7M, and 8M fascicles between segments T2‐T5 observed in the current study can be explained by the great weight of the head and neck in horses (corresponding to 10% of the body mass, according to Buchner et al. 1997; as cited by Zsoldos et al. 2010), which prevents excessive interscapular rotation, principally when the neck is flexed (Denoix, 1999), and is consistent with the increasing relative areas occupied by type I fibres, which are fatigue‐resistant (Table 5).

Nevertheless, we believe that the long multifidus muscle bundles present in the withers (T4 VMS) could act as agonists of the longissimus and spinalis muscles in the cranial portion (T4) of the platform, as suggested by Robert et al. (2001), due to the increased shortening capacity conferred by the increased NFL of the fascicles within these segments (Table 2). These high NFL values are associated with a tendency for decreased PCSA values, indicating a diminished capacity for generating isometric force (Tables 2 and 6). The increased relative areas occupied by type I fibres in these muscles (Tables 4 and 5) can be explained by considering that the architectural design of muscles is a more important predictor of their function than their biochemical characteristics (Lieber & Fridén, 2000; Eng et al. 2008). In addition, type I fibres use energy more efficiently during isometric contractions and slow muscle shortening, according to Bottinelli & Reggiani (2000).

In an evolutionary study conducted in small therian mammals, Schilling (2009) argued that the activity of the longest and most superficial multifidus bundles during asymmetrical gaits (trot and gallop) is appropriately timed to produce column extension.

In contrast, the significantly decreased NFL values and increased PCSA values of 3MT9, 3MT17, 3MT12, 3ML3, and 4ML3 (Tables 2 and 6) increase the isometric muscle force (Felder et al. 2005) and neutralize excessive movements, which are related to the increase in rotation in both the vertical and axial axes between T9‐T14, and the raised transversal rotational axis in T17 (flexion) and lumbosacral region (extension) (Townsend & Leach, 1984; Denoix, 1999; Licka et al. 2009; Groesel et al. 2010).

The R, 3M and 4M muscles act as local stabilizers, global stabilizers and global dynamic stabilizers respectively, according to Schilling (2009). This is evidenced by the PCSA values in the 3MT12, 3MT17, 3ML3, 4MT12 and 4MT17 muscles (Table 6). The aforementioned muscles with high PCSA values consolidate the thoracolumbar platform described by Robert et al. (2001).

When the NFL of the muscles that insert into each of the VMSs were analysed (Table 3), no significant differences between the values for the T9, T12, T17 and L3 VMSs was found. However, the NFL in the T4 VMS was the longest, consistent with its variable mobility according to the neck position (Denoix, 1999), whereas the PCSA values exhibited a significant inverse variation (Table 7). Also of note is the order of the PCSA values for the rotator muscles of RL3, RT4‐RT9 and RT17‐RT12 according to the mobility of the VMS involved (Townsend et al. 1983; Denoix, 1999). The foregoing result suggests that the multifidus muscles and rotator muscles have a heterogeneous antigravitational function (Moseley et al. 2002; Schilling, 2009). With respect to the structure and mechanical action of the rotator muscles and the shorter multifidus muscles (Table 2), Moseley et al. (2002) argued that their polymetameric structure indicates the chronologically different innervation and nonuniform activity of each metamere. The shorter and deeper fascicles could receive nervous stimulation before the more caudal and superficial fascicles do. Our morphological findings are partially consistent with the electromyographic findings obtained in humans (Moseley et al. 2002). The lengths of the multifidus muscles (according to the traverse metamere numbers) appear to be independent of the NFL and muscle mass values (Tables 2 and 8), suggesting their different chronological innervations and diverse functions. Most of the antigravitational capacity observed in the 3M muscles (> PCSA) correspond to their presence in a more mobile VMS. The PCSA values in the L3 VMS are consistent with the high lumbosacral rotational (major extension) capacity of this segment.

Our findings of multifidus bundles with fibres of different lengths packaged within common aponeuroses indicate their adaptive, intersegmental and synergistic capacities (Higham & Biewener, 2011; Meyer & Lieber, 2011) in horses, which allow for better maintenance of intervertebral stability (van Weeren et al. 2010; Stubbs et al. 2011). The fascicles attached to a particular spinous process are supplied by the same spinal nerve (Macintosh et al. 1986); however, this does not mean that they perform a homogeneous function, because each fascicle would behave as part of a neuromuscular compartment (Graziotti et al. 2004). The regionalization and metamerism result in the production of different types of torque, even in a particular joint.

The results obtained in this study suggest that the structural characteristics of the multifidus muscles, NFL, and the functional parameters (PCSA), indicate that they are significant different according to the VMS considered. The results will be helpful as a basis for further research on understanding the posture and dynamics of the equine spine. It can be concluded that the equine multifidus muscle is an immunohistochemically homogeneous muscle with variations in architectural design that have important functional significance, based on the PCSA values.

Acknowledgements

This research was funded by Grant Universidad de Buenos Aires (2014–2017) UBACYT 20020150100102BA Functional anatomy of the equine multifidus muscle with respect to thoracic, lumbar and sacral spinal pathogenesis. The A4.74 monoclonal antibody developed by Helen M. Blau and the BA‐F8 monoclonal antibody developed by Stefano Schiaffino were obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by the Department of Biology at the University of Iowa, Iowa City, Iowa, USA.

References

  1. Barone R (1989) Anatomie Comparée des Mammifères Domestiques, Tome 2: Artrologie et Miologie, pp. 611–615. Paris: Vigot. [Google Scholar]
  2. Bogduk N, Macintosh JE, Percy MJ (1992) A universal model of the lumbar back muscles in the upright position. Spine 17, 897–913. [DOI] [PubMed] [Google Scholar]
  3. Bottinelli R, Reggiani C (2000) Human skeletal muscle fibres: molecular and functional diversity. Prog Biophys Mol Biol 73, 195–262. [DOI] [PubMed] [Google Scholar]
  4. Brooke MM, Kaiser KK (1970) Muscle fibre types: how many and what kind? Arch Neurol 23, 369–379. [DOI] [PubMed] [Google Scholar]
  5. Buchner FH, Savelberg HH, Schamhardt HC, et al. (1997) Inertial properties of Dutch Warmblood Horses. J Biomech 30, 653–658. [DOI] [PubMed] [Google Scholar]
  6. Burkholder TJ, Fingado B, Baron S, et al. (1994) Relationship between muscle fiber types and sizes and muscle architectural properties in the mouse hindlimb. J Morphol 221, 177–190. [DOI] [PubMed] [Google Scholar]
  7. Clayton HM, Kaiser LJ, Lavagnino M, et al. (2012) Evaluation of intersegmental vertebral motion during performance of dynamic mobilization exercises in cervical lateral bending in horses. Am J Vet Res 73, 1–7. [DOI] [PubMed] [Google Scholar]
  8. Delp SL, Suryanarayanan S, Murray WM, et al. (2001) Architecture of the rectus abdominis, quadratus lumborum, and erector spinae. J Biomech 34, 371–375. [DOI] [PubMed] [Google Scholar]
  9. Denoix JM (1999) Spinal biomechanics and functional anatomy. Vet Clin North Am Equine 15, 27–60. [DOI] [PubMed] [Google Scholar]
  10. Dubowitz V, Sewry CA (2007) Histological and histochemical stains and reactions In: A Muscle Biopsy. A Practical Approach (eds Dubowitz V, Sewry CA.), pp. 21–40. London: Saunders Elsevier. [Google Scholar]
  11. Eizema K, van den Burg M, Kiri A, et al. (2003) Differential expression of equine myosin heavy‐chain mRNA and protein isoforms in a limb muscle. J Histochem Cytochem 51, 1207–1216. [DOI] [PubMed] [Google Scholar]
  12. Eng CM, Smallwood LH, Rainiero MP, et al. (2008) Scaling of muscle architecture and fiber types in the rat hindlimb. J Exp Biol 211, 2336–2345. [DOI] [PubMed] [Google Scholar]
  13. Felder A, Ward SR, Lieber RL (2005) Sarcomere length measurement permits high resolution normalization of muscle fiber length in architectural studies. J Exp Biol 208, 3275–3279. [DOI] [PubMed] [Google Scholar]
  14. Ghasemi A, Zahediasl S (2012) Normality tests for statistical analysis: a guide for non‐statisticians. Int J Endocrinol Metab 10, 486–489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Graziotti GH, Palencia P, Delhon G, et al. (2004) Neuromuscular partitioning, architectural design and myosin fiber types of the M. vastus lateralis of the llama (Lama glama). J Morphol 262, 667–681. [DOI] [PubMed] [Google Scholar]
  16. Groesel M, Zsoldos RR, Kotschwar A, et al. (2010) A preliminary model study of the equine back including activity of longissimus dorsi muscle. Equine Vet J 42(Suppl. 38), 401–406. [DOI] [PubMed] [Google Scholar]
  17. Hansen L, de Zee M, Rasmussen J, et al. (2006) Anatomy and biomechanics of the back muscle in the lumbar spine with reference to biomechanical modeling. Spine 31, 1888–1899. [DOI] [PubMed] [Google Scholar]
  18. Haussler KK (1999) Anatomy of the thoracolumbar vertebral region. Vet Clin North Am Equine 15, 13–25. [DOI] [PubMed] [Google Scholar]
  19. Hermanson JW (1997) Architecture and the division of labor in the extensor carpi radialis muscle of horse. Acta Anat 159, 127–135. [DOI] [PubMed] [Google Scholar]
  20. Higham TE, Biewener AA (2011) Functional and architectural complexity within and between muscles: regional variation and intermuscular force transmission. Philos Trans R Soc B 366, 1477–1487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hyytiäinen HK, Mykkänen AK, Hielm‐Björkman AK, et al. (2014) Muscle fibre type distribution of the thoracolumbar and hindlimb regions of horses: relating fibre type and functional role. Acta Vet Scand 56, 1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kaigle AM, Sten MS, Holm H, et al. (1995) Experimental instability in the lumbar spine. Spine 20, 421–430. [DOI] [PubMed] [Google Scholar]
  23. Kearns CF, McKeever KH, Abe T (2002) Overview of horse body composition and muscle architecture: implications for performance. Vet J 164, 224–234. [DOI] [PubMed] [Google Scholar]
  24. Licka T, Frey A, Peham C (2009) Electromyographic activity of the longissimus dorsi muscles in horses when walking on a treadmill. Vet J 180, 71–76. [DOI] [PubMed] [Google Scholar]
  25. Lieber RL, Fridén J (2000) Functional and clinical significance of skeletal muscle architecture. Muscle Nerve 23, 1647–1666. [DOI] [PubMed] [Google Scholar]
  26. Macintosh JE, Bogduk N (1986) The biomechanics of the lumbar multifidus. Clin Biomech 1, 206–213. [DOI] [PubMed] [Google Scholar]
  27. Macintosh JE, Valencia F, Bogduk N, et al. (1986) The morphology of the human lumbar multifidus. Clin Biomech 1, 196–204. [DOI] [PubMed] [Google Scholar]
  28. Marx JO, Olsson MC, Larsson L (2006) Scaling of skeletal muscle shortening velocity in mammals representing a 100,000‐fold difference in body size. Pflugers Arch 452, 222–230. [DOI] [PubMed] [Google Scholar]
  29. Mc Gowan C, Stubbs N, Hodges P, et al. (2007) Back Pain in Horses, pp. 9–35. Barton, ACT: Rural Industries Research and Development Corporation. Australian Government ISSN 1440‐6845. [Google Scholar]
  30. Meyer GA, Lieber RL (2011) Elucidation of extracellular matrix mechanics from muscle fibers and fiber bundles. J Biomech 44, 771–773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Moore KK, Dalley AF, Agur AMR (2013) Anatomía con Orientación Clínica. Philadelphia: Wolters Kluwer/Lippincott Williams & Wilkins. [Google Scholar]
  32. Moseley GL, Hodges PW, Gandevia S (2002) Deep and superficial fibers of the lumbar multifidus muscle are differentially active during voluntary arm movements. Spine 27, E29–E36. [DOI] [PubMed] [Google Scholar]
  33. Nickel RA, Schummer E, Seiferle H, et al. (1985) The Anatomy of the Domestic Animals, Volume 1: The Locomotor System of the Domestic Mammals, pp. 293–294. Berlin: Verlag Paul Parey. [Google Scholar]
  34. Nwoye L, Mommaerts WFHM, Simpson DR, et al. (1982) Evidence for a direct action of thyroid hormone in specifying muscle properties. Am J Physiol 242, R401–R408. [DOI] [PubMed] [Google Scholar]
  35. Panjabi M. M. (1992) The stabilizing system of the spine. Part I. Function, dysfunction, adaptation, and enhancement. J. Spinal Disord 5, 383–389. [DOI] [PubMed] [Google Scholar]
  36. Payne RC, Veenmam P, Wilson AM (2005) The role of the extrinsic thoracic limb muscles in equine locomotion. J Anat 206, 193–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ritter DA, Nassar PN, Fife M, et al. (2001) Epaxial muscle function in trotting dogs. J Exp Biol 204, 3053–3064. [DOI] [PubMed] [Google Scholar]
  38. Rivero JLL, Talmadge RJ, Edgerton R (1996) Myosin heavy chain isoforms in adult equine skeletal muscle: an immunohistochemical and electrophoretic study. Anat Rec 246, 185–194. [DOI] [PubMed] [Google Scholar]
  39. Robert C, Audigié F, Valette JP, et al. (2001) Effects on treadmill speed on the mechanics of the back in the trotting saddlehorse. Equine Vet J 33(Suppl.), 154–159. [DOI] [PubMed] [Google Scholar]
  40. Rosatelli A. L., Ravichandiran K. and Agur A. M. (2008) Three‐dimensional study of the musculotendinous architecture of lumbar multífidus and its functional implications. Clin. Anat 21, 539–546. [DOI] [PubMed] [Google Scholar]
  41. Schilling N (2009) Metabolic profile of the perivertebral muscles in small therian mammals: implications for the evolution of the mammalian trunk musculature. Zoology 112, 279–304. [DOI] [PubMed] [Google Scholar]
  42. Schilling N, Carrier DR (2009) Function of the epaxial muscles during trotting. J Exp Biol 212, 1053–1063. [DOI] [PubMed] [Google Scholar]
  43. Sharir A, Milgram J, Shahar R (2006) Structural and functional anatomy of the neck musculature of the dog (Canis familiaris). J Anat 208, 331–351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Stubbs NC, Hodges PW, Jeffcott LB, et al. (2006) Functional anatomy of the caudal thoracolumbar and lumbosacral spine in the horse. Equine Vet J 38(Suppl. 36), 393–399. [DOI] [PubMed] [Google Scholar]
  45. Stubbs NC, Kaiser LJ, Hauptman J, et al. (2011) Dynamic mobilisation exercises increase cross sectional area of musculus multifidus. Equine Vet J 43, 522–529. [DOI] [PubMed] [Google Scholar]
  46. Townsend HGG, Leach DH (1984) Relation between intervertebral joint morphology and mobility in the equine thoracolumbar spine. Equine Vet J 16, 461–465. [DOI] [PubMed] [Google Scholar]
  47. Townsend HGG, Leach DH, Fretz PB (1983) Kinematics of the equine thoracolumbar spine. Equine Vet J 15, 117–122. [DOI] [PubMed] [Google Scholar]
  48. van Weeren PR (2009) Kinematics of the equine back. Equine back pathology diagnosis and treatment, capítulo 5 Editor Henson FMD. Wiley‐ Blackwell, Singapur. [Google Scholar]
  49. van Weeren PR, Mc Gowan C, Haussler KK (2010) Developmental of a structural and functional understanding of the equine back. Equine Vet J 42(Suppl. 38), 393–400.20636774 [Google Scholar]
  50. Zaneb H, Peham C, Stanek C (2013) Functional anatomy and biomechanics of the equine thoracolumbar spine. A review. Turk J Vet Anim Sci 37, 380–389. [Google Scholar]
  51. Zsoldos RR, Kotschwar AB, Kotschwar A, et al. (2010) Electromyography activity of the equine splenius muscle and neck kinematics during walk and trot on the treadmill. Equine Vet J 42(Suppl. 38), 455–461. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Anatomy are provided here courtesy of Anatomical Society of Great Britain and Ireland

RESOURCES